Schedule Crashing Calculator
This schedule crashing calculator finds the lowest additional direct cost that recovers a requested number of project days in a small predecessor network. It searches whole-day activity durations and recalculates the complete schedule for each candidate, so parallel critical paths are handled together.
By Dr. Hassan Khames EliwaLast updated
Result
Exact within the stated whole-day, linear-cost model; excludes indirect costs and resource constraints.
- Normal project duration
- 8 days
- Target duration
- 6 days
- Selected project duration
- 6 days
- Selected total direct cost
- $1,650.00
- Selected critical pathsAll critical paths shown.
- A → B → D; A → C → D
One minimum-cost feasible plan; equivalent plans may exist
| Activity | Normal days | Selected days | Days crashed | Cost per day saved | Added cost |
|---|---|---|---|---|---|
| A | 3 | 2 | 1 | $200.00 | $200.00 |
| B | 4 | 3 | 1 | $100.00 | $100.00 |
| C | 4 | 3 | 1 | $150.00 | $150.00 |
| D | 1 | 1 | 0 | $0.00 | $0.00 |
How it works
Describe the network before buying time
Each activity needs an ID, normal duration, crash duration, normal cost, crash cost, and predecessors. Durations are whole days. Predecessors use finish-to-start relationships with zero lag; separate multiple predecessors with spaces or a vertical bar. Use a dash for an activity with no predecessors.
Normal duration is the unaccelerated estimate. Crash duration is the shortest permitted duration, not the number of days you want to remove. Costs are total direct activity costs at those endpoints. IDs must be unique, references must exist, and the network cannot contain a cycle.
Price each recoverable activity day
The assumed cost slope is (crash cost − normal cost) ÷ (normal duration − crash duration). Intermediate whole-day durations are priced by linear interpolation. An activity going from four days at $400 to two days at $600 therefore costs $100 for each day removed.
Equal duration endpoints must have equal costs because there is no duration change to price. The model accepts zero-day milestones and zero-cost acceleration. It rejects declining crash costs, since the search and its pruning assume acceleration never reduces direct cost.
Exact search rather than a critical-path shortcut
The calculator first schedules the normal network and subtracts your requested recovery to establish a target. It then checks the all-crashed network for feasibility. Within the accepted bounds, it enumerates permitted duration combinations, recomputing early dates, late dates, float, and critical paths for candidate plans. Branches already more expensive than a feasible incumbent can be discarded safely.
A plan qualifies only when its entire network finishes by the target. This matters when two paths are critical: shortening only one usually buys no project recovery. A shared activity might shorten both paths, while separate activities may need to be crashed as a package.
The result is an exact minimum within this discrete linear-cost model, not a greedy claim of global optimality for real projects. Inputs are limited to 12 activities, normal durations of at most 20 days, and 10,000 duration combinations. Larger problems are rejected rather than answered with an unlabeled approximation. GAO’s Schedule Assessment Guide provides the broader context for reliable logic and critical-path analysis.
Worked example
Two parallel critical branches
The default network starts with A, then runs B and C in parallel, and finishes with D after both branches. Their normal durations are three, four, four, and one days, so the project takes eight days. Normal direct cost totals $1,200.
A can lose one day for $200. B can lose two days for $100 per day. C can lose two days for $150 per day. D cannot be shortened. To recover two project days, crashing B alone is ineffective because C still controls the finish.
The minimum-cost plan shortens A by one day, B by one, and C by one. Added cost is $200 + $100 + $150 = $450. Both A–B–D and A–C–D take six days. Selected total direct cost is $1,650. Shortening both B and C by two days would also finish in six days, but costs $500.
A single recoverable chain
For A at four days and $400, crashable to two days at $600, followed by fixed B at two days and $200, normal duration is six days. Recovering two days costs $200 and produces a four-day schedule with $800 total direct cost. Requesting three days is infeasible.
Expert tips
Common mistakes
- Crashing only the original longest path. Once durations change, another path may become critical; this calculator reschedules the entire network instead of assuming the original path remains controlling.
- Entering incremental crash premiums as crash costs. Enter the complete activity cost at the shortest duration; the calculator subtracts normal cost itself.
- Assuming the cheapest displayed plan is unique. Tied allocations can exist. The output provides one minimum-cost feasible plan rather than listing every equivalent allocation.
Frequently asked questions
Does this use a greedy crashing rule?
No. It uses bounded exact enumeration with cost pruning. A cheapest-activity greedy rule can fail when parallel critical paths require coordinated reductions.
Can activities overlap?
Activities without a dependency may run in parallel. The model does not support start-to-start links, leads, lags, or partial overlap within a dependency.
Why reject a large network?
The number of combinations multiplies across crash ranges. A hard 10,000-combination limit keeps browser work bounded and preserves the exact-solution promise for accepted inputs.
Can I recover zero days?
Yes. The normal schedule meets that target with zero added cost. This is useful for checking the network and its baseline direct cost before testing recovery options.
Are fractional-day options supported?
No. Durations and requested recovery are whole days. Fractional durations or nonlinear overtime packages require a different optimization model and may have a different least-cost solution.
Related calculators
Related guides
Construction & ProjectThree-point estimating: when PERT beats a single guessA practical guide to three-point duration estimates, weighted averages and the limits of turning uncertain activities into a promised date.14 November 2026 · 7 min read
Construction & ProjectEarned value management in plain English: CPI, SPI and what to do about themTurn planned value, earned value and actual cost into useful decisions without confusing budget performance with cash flow or time.14 November 2026 · 7 min readSources and standards
About the author
Dr. Hassan Khames Eliwa holds a PhD in Construction & Built Environment and has more than 20 years of international project controls experience. He is Director of Eliwa Project Controls Limited in Auckland and founder of PMMilestone.org and PMMilestone.com.
Buy me a coffee