The Recovery Schedule That Created a New Bottleneck
A 10-month hospital fit-out is 3 months behind at month 7. The recovery plan compresses the remaining mechanical, electrical, and finishes work by overlapping trades and adding a second shift. The scheduler produces a recovery schedule showing completion back on the original date. Resource loading shows: electrical crew demand peaks at 40 workers in weeks 3–5 of the recovery, but the site can support only 25 at once; the second shift requires a safety orientation that takes 5 days per new worker; and the mechanical rough-in cannot be inspected until the electrical rough-in in the same zones is complete (a code requirement). The project manager wants to issue the recovery schedule immediately to satisfy the client.
1. Evaluate whether the recovery schedule is achievable. What defects must be corrected before issue?
Read the worked answer
The recovery schedule is not achievable as drawn. Three defects:
Resource over-allocation. Electrical demand peaks at 40 workers but the site supports 25. An unleveled schedule will fail in execution. The plan must resource-level the electrical work (spread it, sequence zones, or reduce concurrent scope) and re-run the schedule to see the true duration. The 15-worker excess is a hard constraint, not a preference.
Second-shift ramp-up not modeled. A 5-day safety orientation per new worker is real duration. If new workers are needed to hit the peak, the orientation must be an activity with predecessors and successors-otherwise the second shift "starts" before workers are actually cleared, and the schedule is fiction.
Inspection logic ignored. Mechanical rough-in cannot be inspected until electrical rough-in in the same zone is complete (code). If the recovery overlaps mechanical ahead of electrical, the mechanical work will be held at inspection-creating rework or idle time. The overlap must respect the inspection sequence, zone by zone.
Before issue: level resources, model orientation and inspection logic, re-run the critical path, and confirm the recovery date is still achievable. If it isn't, report the realistic date and the gap-do not issue a schedule that cannot be executed.
2. The PM insists the client needs a date today. What should the scheduler provide, and how should the report frame the recovery?
Read the worked answer
Provide a defensible date, not a wished-for one.
- Issue the resource-leveled, logic-corrected schedule with the realistic forecast completion date, and show the gap between the original date and the forecast date.
- Present the recovery as a range or scenario set: e.g., "Base recovery (leveled): date X; with added crew capacity (site expansion): date Y; with utility/inspection resequencing: date Z." Give the client options with cost and risk.
- Show the binding constraints explicitly: 25-worker site cap, 5-day orientation, inspection sequence. These are the levers-if the client can expand site capacity or accept phased turnover, the date improves.
- Include a risk register for the recovery: second-shift productivity, inspection delays, worker availability.
Framing: "Here is the earliest achievable date given the constraints, here is what would have to change to hit the original date, and here is the cost/risk of each option." That is more useful-and more credible-than an unachievable date issued to satisfy a meeting.
Judgment point: the exam tests whether you refuse to issue a schedule that violates resource, safety, or inspection logic, and whether you convert a binary "can we hit the date?" into a structured options analysis.