Earned Value, Weather Delay, and Whether to Buy Acceleration

Arithmetic

• CPI = 880,000 / 956,500 = 0.920
• SPI = 880,000 / 1,000,000 = 0.880
• CV = 880,000 - 956,500 = -$76,500
• SV = 880,000 - 1,000,000 = -$120,000 of planned work not yet placed
• EAC = 2,400,000 / 0.920 = $2,608,700
• VAC = 2,400,000 - 2,608,700 = -$208,700

What it does and does not mean. Both indices are below 1.0 and both are roughly in the same range, which points to a systemic productivity or buyout problem rather than a single bad package - but the BAC/CPI forecast assumes the inefficiency continues at the same rate for eight more months, which is a strong assumption. Do not present the $2.6 million figure as a verdict. Before Monday, break the variances down by cost code and by trade: if the overrun sits in one package (a bad buyout, a rework event), a bottom-up forecast will be materially better than EAC. If it is spread evenly across labor codes, the CPI forecast is probably optimistic, because the schedule pressure will make the productivity problem worse, not better. Present it as a range with the assumption stated.

Grant six days of time and zero dollars.

Reasoning, step by step:

1. Excusable but not compensable. Weather of the type the contract contemplates is an excusable delay - it buys time, not money. No cost, no extended general conditions, no winter-condition markup.
2. Float belongs to the project. The baseline shows three days of total float on that activity. A nine-day delay consumes the three days of float and then pushes the contract completion by the remainder: 9 - 3 = 6 days. Float is not for the owner's or the sub's exclusive use; what it does is define when a delay becomes a completion delay.
3. Test the facts before you sign anything. Confirm the days were actually lost (daily reports, weather records against the project calendar), confirm whether the contract is calendar-day or working-day, and confirm the float is on the path the sub is actually driving and has not already been consumed by another delay.
4. Watch concurrency. If other work was already delaying completion on those same days, granting the six days does not reduce your own exposure - but refusing it does not help you either, and a refusal you cannot defend invites a claim.

Give the six days by formal change order with a revised completion date and a written denial of cost in the same document. If you refuse outright, the sub's entitlement does not disappear; it simply moves to a forum where you also have to explain your own five-week slip.

No - not as a blanket purchase.

• Five weeks of exposure = 35 calendar days x $2,500 = $87,500 in potential liquidated damages.
• The acceleration package costs $95,000.
• Spending $95,000 to avoid $87,500 is a $7,500 net loss even in the best case where the six weeks are fully recovered - and full recovery from overtime is the assumption most often wrong, because overtime on a congested site buys less than a linear time gain.
• The LD cap is 10 percent of the contract price, $240,000, so the exposure is bounded, and LDs are only assessed for delay the contractor is responsible for. Weather delay is excusable; if a meaningful part of the five weeks is excusable, the realistic LD exposure is smaller than $87,500 before you spend a dollar.

What to do instead. Buy acceleration selectively where the marginal cost of a recovered day is under $2,500. The electrical package that just got six days is the obvious target: a few extra electricians on the critical path may recover those six days for a few thousand dollars, which is a far better trade than overtime across the whole job. Present the owner with the arithmetic, not a promise: here is our forecast, here is what we can recover at what cost, and here is what we recommend. Owners respond better to a cost-per-day comparison than to a heroic acceleration offer that shows up later as a change order.