Glossary

The number of days an activity can slip before it pushes the project's completion date. It is a property of the network, not of the activity, and it changes every time the programme is updated.

The number of days an activity can slip without delaying any activity that immediately follows it. It is always equal to or smaller than total float. Consuming free float is invisible at project level, which is why it is often consumed without anyone noticing.

The sequence of activities that determines the project's completion date, conventionally the one with zero or least total float. A delay to any activity on it delays the project unless the sequence is changed. In a real programme there is usually more than one, and they move between updates.

A sequence with a small amount of float: enough to be ignored in reporting, not enough to survive a normal month. Projects are routinely delayed by paths that were never on the critical path and were therefore never managed.

The continuous chain of activities driving the finish date, traced backwards from completion. It is used in place of the zero-float definition of criticality when calendars, constraints or lags distort float values.

The agreed schedule against which progress and delay are measured, fixed at a point in time and formally accepted. Everything in delay analysis is measured against it, which is why a baseline that was never properly accepted undermines every later argument.

Two delays affecting the same period, one the employer's responsibility and one the contractor's. It usually determines whether the contractor gets time, money, both or neither.

A deliberate slowing of work in response to another delay, on the reasoning that there is no point rushing to a gate that is already closed. It is claimed as a reasoned management decision rather than a failure.

A delay for which the contractor is entitled to more time, because its cause is not their responsibility under the contract. It protects against liquidated damages but does not by itself carry money.

An excusable delay that also carries an entitlement to cost, because its cause is the employer's responsibility. Every compensable delay is excusable; the reverse is not true, and confusing the two is the most common error in a first claim.

A prospective method that inserts a delay event into the programme as it stood immediately before the event, and measures the movement in the completion date. It answers what the delay was expected to do, using information available at the time.

A retrospective comparison of the programme as accepted against the works as actually carried out. It is the most intuitive method and the least rigorous, because it shows what happened without establishing what caused it.

Working faster or with more resources to recover time or finish early, usually at additional cost. Instructed acceleration is agreed and paid for; the argument is about the uninstructed kind.

A lag delays a successor activity by a stated period after its predecessor's link condition is met; a lead pulls it forward. They are modelling conveniences, and large lags often hide a real activity nobody wanted to draw.

Shortening a programme either by adding resource to critical activities, or by overlapping activities that were planned in sequence. The first buys time with money; the second buys it with risk.

An instruction in the programme that fixes an activity to a date rather than letting the logic determine it. Constraints suppress the network's own calculation, which is why a heavily constrained programme can report float that does not exist.

A formal extension of the contractual completion date, granted for a delay the contract says is not the contractor's risk. Its first function is to protect the contractor from liquidated damages; its second is to keep the completion date meaningful.

A sum stated in the contract, payable per day or per week of late completion, agreed in advance instead of proving actual loss. It is a cap and a certainty as much as it is a threat.

The time-related cost of remaining on site longer than planned: site establishment, supervision, plant standing, accommodation. It compensates the period of delay, not the work done in it.

Loss of productivity, meaning the same work taking more labour hours than it should have, caused by interference, changed sequence, congestion or late information. It can occur without any delay to the completion date at all.

A claim that attributes a total cost overrun to a group of events without linking each event to a specific part of the loss. It is made when the records will not support event-by-event causation.

An instructed change to the scope, quality or sequence of the works, valued under the contract's own rules. Whether something is a variation or was always included is where most commercial argument on a project actually happens.

A formal, contractually required communication that an event has occurred or a claim will be made, usually within a stated period. Its purpose is to give the other party the chance to investigate and mitigate while it still can.

A contractual mechanism excusing performance when a defined exceptional event prevents it. In common-law contracts it exists only if the contract creates it; in many civil-law systems it exists in the code whether the contract mentions it or not.

The position where there is no longer an enforceable completion date, so the obligation becomes to finish within a reasonable time. It typically arises where the employer has prevented completion and the contract's extension mechanism has broken down. Liquidated damages generally fall away with it.

A percentage of each payment withheld by the employer as security for defects and completion, released in stages. It is the contractor's money held against future performance, and its release schedule is a real cash-flow item.

A bank or insurer's undertaking to pay the employer a stated sum if the contractor fails to perform. On-demand bonds pay against a compliant demand alone; conditional bonds require proof of default. The difference is enormous and is decided by the bond wording, not by its title.

A bank guarantee securing an advance payment, reducing as the advance is recovered through certified work. It is what makes a large advance commercially acceptable to the paying party.

Subcontract terms drafted to mirror the main contract, so that obligations and risks pass down intact. They fail in the details: notice periods that are impossible to meet once a tier is added, and payment terms that depend on someone else being paid first.

An allowance included in the contract price for work that is not yet defined, to be substituted by a proper valuation once it is. It is a placeholder, not a budget, and treating it as a budget is how scope arguments start.

A set of three-letter trade terms published by the International Chamber of Commerce that allocate delivery, cost and risk between seller and buyer. They do not deal with title, payment or governing law. The rule is only half the term. The named place beside it does the other half.

The full cost of a unit delivered to your door: goods, freight, insurance, duty, non-recoverable taxes, handling, finance and any amortised one-off charges. It is the number that should be compared between suppliers, and almost never is.

The weight an airline actually bills: the greater of actual gross weight and volumetric weight, where volume in cubic metres is converted at a standard factor. A light, bulky shipment pays for the space it occupies rather than the mass it has.

The ocean-freight convention of billing on the greater of a shipment's tonnes and its cubic metres, where one revenue tonne is one tonne or one cubic metre, whichever is more. It exists for the same reason as air's volumetric weight: space on a vessel is finite in both dimensions.

A carrier's document that receipts the goods, evidences the contract of carriage and, in its negotiable form, controls delivery of the cargo. A negotiable bill is a document of title, which is why originals are couriered and why losing one is a serious event.

A bank's undertaking to pay the seller against documents that comply exactly with stated terms. Banks deal in documents, not in goods, so a compliant presentation is paid whether or not the shipment is what was expected.

A charge for keeping cargo or equipment at the terminal beyond the free time allowed. In container trade it is generally the charge for the box sitting inside the port.

A charge for keeping the carrier's equipment beyond the free time once it has left the terminal, which is the container standing at your yard while it is unpacked. Free time on this and on demurrage are separately negotiated and separately consumed.

Assigning goods to a heading in the harmonised tariff, which determines the duty rate, the permits required and the restrictions that apply. It is a legal determination about what a thing is, and neighbouring headings can differ by many percentage points.

The party legally responsible to the customs authority for the declaration, the duty and anything reassessed later. It is a registered role in the destination country, and it is not always the party that receives the goods.

A document attesting that goods meet a stated standard or regulation, often required before customs release in Gulf markets. It is issued against a specific consignment or model, and a certificate that does not match the shipment is worth nothing at the border.

The elapsed time between placing an order and having the goods available to use. Every part of that sentence is negotiable in practice, which is why two people quoting a lead time are often quoting different things.

Actively chasing an order through a supplier's process to protect its date, rather than waiting for a status report. Done well it is early detection; done badly it is a queue of calls that changes nothing and teaches suppliers to give reassuring answers.

Stock held to absorb variation in demand and in lead time, above the quantity needed to cover expected consumption. It is bought with the variability of the supply, not with its average.

The stock level at which a replenishment order is placed: expected demand over the lead time, plus safety stock. If the lead time moves and the reorder point does not, the safety stock quietly stops being safety stock.

The share of demand a stocking policy intends to satisfy from stock. It is a policy decision that sets how much safety stock is carried, not a measurement of what happened.

The proportion of demand satisfied from available stock. It is one of the most quoted numbers in supply chain and one of the least defined.

A combined measure of whether an order arrived on the agreed date and complete. Both halves have to be true for it to count.

How many times inventory is sold and replaced over a period, used as a measure of how hard working capital is working. Reported alone, it says less than it appears to.

Stock that has been paid for or committed and is in transit, not yet available to use. It is funded exactly like stock on the shelf, and it grows in direct proportion to lead time.

The amplification of demand variability as it moves up a supply chain, so that a small change in end demand produces a large swing in factory orders. Batching, forecast updating, price promotions and shortage gaming each contribute.

Stock on hand expressed as the number of days of expected demand it would satisfy. It is easier to reason about than a unit count because it is directly comparable to a lead time.

The smallest quantity a supplier will accept for an order, set by set-up cost, batch size or commercial policy. It converts a demand problem into an inventory problem, and the inventory it creates is rarely priced into the unit cost comparison.

A consistent tendency for forecasts to run above or below actual demand, as distinct from the size of the error. Bias is correctable and error is not, which is why measuring only accuracy hides the fixable half of the problem.