Build a precedence network for a newbuild project, find the critical path by hand, and use float to protect the schedule that actually matters.
The critical path is the single longest chain of dependent activities in your schedule — and on a newbuild project, it is rarely where people first assume it is. This lesson walks through building a precedence network, calculating the critical path by hand, and using float correctly so you protect the dates that actually determine delivery.
Why CPM Matters More on a Newbuild Than on Most Projects
A shipbuilding schedule routinely carries several thousand activities across engineering, procurement, steel, outfitting and trials — with long-lead equipment (main engines, gearboxes, generators) ordered a year or more before it is installed, and hundreds of block-level activities that all compete for the same dock and crane capacity. Without a disciplined critical path, planners default to tracking "everything," which means nothing gets the attention it needs. CPM gives you a defensible answer to the only question that matters: if this activity slips by a week, does the delivery date slip too?
Building the Precedence Network
Every activity in a CPM network has a duration and one or more predecessors. The four relationship types you'll actually use:
- Finish-to-Start (FS): the default — a successor can't start until its predecessor finishes. "Steel Cutting" must finish before "Block Fabrication" can start.
- Start-to-Start (SS): the successor can start once the predecessor starts, often with a lag. "Piping Installation" can start 2 weeks after "Block Assembly" starts, without waiting for it to finish.
- Finish-to-Finish (FF): both activities must finish together, or one within a fixed offset of the other. Useful for coating and NDT sign-off that have to close out in lockstep.
- Start-to-Finish (SF): rare in practice — mostly seen in shift-handover or just-in-time delivery logic. You can usually avoid it.
Resist the temptation to make everything Finish-to-Start "to keep it simple." Overusing FS is the single most common reason a shipyard schedule shows a fake critical path that doesn't match how the yard actually works — piping, cabling and insulation genuinely do overlap block assembly, and modelling that overlap with SS relationships and a realistic lag is what makes the network trustworthy.
A Worked Example
Take a simplified slice of a newbuild network for one hull section:
- A — Steel Cutting: 10 days, no predecessor
- B — Sub-block Fabrication: 20 days, starts after A finishes (FS)
- C — Block Assembly: 15 days, starts after B finishes (FS)
- D — Pre-erection Outfitting: 12 days, starts 5 days after C starts (SS+5)
- E — Erection: 8 days, starts after C and D both finish (FS from both)
Running a forward pass (earliest start/finish) from day 0: A finishes day 10; B finishes day 30; C finishes day 45; D starts day 20 (5 days after C starts on day 30 — note the SS lag is measured from C's start, so D actually starts day 35) and runs 12 days to day 47; E needs both C (day 45) and D (day 47) finished, so E starts day 47 and finishes day 55. The path A→B→C→D→E is critical at 55 days — notice that D, not C, ends up driving E's start, even though C looked like the "main line" of the network. This is exactly the kind of result that only shows up once you calculate the network properly instead of eyeballing it.
Total Float vs. Free Float
Total float is how much an activity can slip without delaying the project finish date. Free float is how much it can slip without delaying its immediate successor. The distinction matters operationally: a production foreman only cares about free float ("can I move this crew tomorrow without blocking the next crew"), while the project manager needs total float to know which near-critical chains deserve a buffer. A common mistake is treating every activity with float as "safe" — a chain of activities each with 2 days of float, sharing the same float pool, can consume it all collectively even though no single activity looks urgent. Always check float at the path level, not just the activity level.
Where Shipyard Schedules Diverge from Textbook CPM
Pure CPM assumes unlimited resources — in a shipyard that assumption breaks immediately. The dock crane, the dry dock itself, and skilled welder headcount are all hard constraints that can make a logically non-critical activity into the real bottleneck, simply because it's competing for the same crane slot as something on the critical path. This is exactly the subject of the next lesson in this series, Resource Leveling and Crane/Dock Capacity Planning — treat the two as a pair: CPM tells you the logical sequence, resource leveling tells you whether the yard can actually execute it in that order.
Practical Tips for Newbuild Planners
- Baseline early and don't quietly edit it. Track slippage against the original baseline, not a constantly-shifting "current plan" — otherwise you lose the ability to see how much total delay has actually accumulated.
- Update on a fixed cadence (weekly is typical for production, monthly for the overall program) so the critical path calculation stays meaningful instead of being recalculated ad hoc every time someone asks.
- Avoid hard date constraints ("must finish by") on individual activities unless they represent a real external commitment — they silently break the float calculation and can hide a true critical path behind an artificial one.
- Review the critical path itself at every update, not just percent-complete. The path frequently moves from one system to another (e.g. from structural steel early in the project to electrical commissioning near the end) and planners who don't re-check it keep managing yesterday's bottleneck.
Our free Newbuild Master Schedule Template is a ready-to-edit starting point with the WBS, dependencies and milestone structure already laid out — a practical way to put this lesson's network logic straight into a working schedule.
