Project Controls Q&A — 120+ professional answers

The project plan is the integrated document covering scope, schedule, cost, quality, risk, communications and stakeholders. The schedule is one component of that plan — the time-phased model of when activities occur and how they depend on each other.

Decompose the scope deliverable-by-deliverable, not activity-by-activity. Stop at the level where a single owner can plan, estimate and report on the package weekly. Every WBS element should describe a tangible outcome, not a task verb.

Rolling-wave planning details near-term work in fine grain and leaves longer-term work at a higher WBS level until information is available. Use it on projects with discovery, design evolution, or long horizons such as research, IT transformation and mega-construction.

It is the practice of refining estimates, scope, schedule and risk as the team learns more. It is the antidote to false precision — you commit only what is currently knowable and tighten the plan in subsequent baselines.

Without a baseline you cannot calculate variance, you cannot earn value, and you cannot defend an extension of time or a cost overrun. The baseline is the contractually and analytically defensible reference point.

Typical industry guidance keeps activities between 1 and 20 working days. Smaller activities exaggerate progress; larger ones hide drift. Tune the duration band to your reporting cycle and the discipline involved.

Finish-to-Start (FS), Start-to-Start (SS), Finish-to-Finish (FF) and Start-to-Finish (SF). FS dominates real schedules; SS and FF are common in construction overlaps; SF is rare and should be challenged whenever it appears.

Lags are justified for curing, dewatering, settlement, procurement lead times and similar physical or commercial waits. They are not a substitute for missing activities — if a lag exceeds about 10 working days, model the underlying work explicitly.

A control account is a management point where scope, schedule and cost converge. It is where EVM is calculated and where a single accountable manager owns performance. Without control accounts, EVM degenerates into project-level averages.

Use three-point PERT estimates ((O+4M+P)/6) on uncertain activities, anchor mature activities to historical productivity data, and apply parametric estimates where quantities and rates are known. Document the basis of estimate for every long-duration activity.

Effort is total person-hours required. Duration is calendar time across which the work is performed. Doubling resources may halve duration up to a point but rarely halves effort due to coordination overhead.

Missing logic, hidden lags, over-constraining dates, omitted scope, unrealistic productivity assumptions and not aligning the schedule to the contract milestones. Most schedule failures are planning failures, not execution failures.

The longest sequence of dependent activities through the schedule network. It defines the shortest possible project duration; any delay on a critical activity delays project completion unless mitigated.

Yes. Whenever multiple paths share the same longest duration, all are critical. On complex projects with many concurrent fronts, near-critical paths (within 5–10 days of the critical path) require equal management attention.

A set of 14 schedule quality checks developed by the US Defense Contract Management Agency. They evaluate logic, leads, lags, relationship types, hard constraints, high float, negative float, high duration, invalid dates, resources, missed tasks, critical path test, CPLI and BEI.

Most quality frameworks expect 95–100% of activities to have at least one predecessor and one successor, with less than 5% missing logic. Open-ended activities distort the critical path and corrupt forecast dates.

Hard constraints (Must Finish On, Start No Later Than) override CPM logic and hide slippage. They produce negative float without showing the underlying driver. Limit them to genuine contractual milestones.

Leveling adjusts dates to resolve resource overloads and will extend the schedule if needed. Smoothing redistributes work only within available float and never extends the schedule.

Industry standard is weekly status updates with a formal monthly progress update tied to invoicing. The contract usually specifies the data date cadence; deviating from it weakens any future claim.

Techniques used to shorten a schedule without reducing scope, primarily crashing (adding resources to critical activities) and fast-tracking (running activities in parallel that were planned sequentially). Both carry cost or rework risk.

Fast-tracking executes activities in parallel that were sequenced FS in the baseline. It is dangerous when downstream work depends on still-evolving deliverables — rework probability climbs sharply with the degree of overlap.

Adding resources to critical activities to shorten their duration. It only works on activities where duration is truly resource-driven, not on serial activities like curing, inspection or commissioning sequences.

Use physical percent complete for measurable work, weighted milestones for phased deliverables, units complete for repetitive work, and 0/50/100 only for short-duration activities. Avoid level-of-effort except for support functions.

A short-term schedule (typically 3 to 6 weeks) extracted from the master schedule and used by site teams to coordinate near-term work, identify constraints and confirm resource availability.

A revised schedule produced after material delay that demonstrates how the contractor will recover lost time and finish by the contractual completion date, including any acceleration or sequence change required.

Only via approved schedule revision following a contractual mechanism — usually a major change in scope, an extension of time award, or acceleration directive. Casual re-baselining destroys variance integrity.

Linking schedule activities to BIM model elements so progress, sequence and clash detection can be visualised in time. It is now standard on mega infrastructure and complex industrial builds.

It collapses scope, time and cost into a single objective measure of performance so a project manager can answer, in one number, 'how is the project really doing today?' without relying on opinion.

SPI = EV ÷ PV. A value of 1.0 is on plan; 0.9–1.0 is generally acceptable; below 0.85 demands a recovery plan. Above 1.05 may indicate front-loaded progress claims rather than genuine over-performance.

CPI exposes cost efficiency — whether you are getting one dollar of work for every dollar spent. SPI can look healthy while CPI is collapsing, which usually means accelerated spend has bought temporary schedule progress.

Use EAC = BAC ÷ CPI when current performance is expected to continue, EAC = AC + (BAC − EV) when the variance is a one-off, and EAC = AC + (BAC − EV)/(CPI × SPI) when both cost and schedule pressure are sustained.

Earned Schedule reports schedule performance in time units rather than dollars. SPI degrades as the project nears completion (EV → BAC, PV → BAC, ratio → 1.0). SPI(t) does not suffer this distortion.

TCPI is the cost performance the remaining work must achieve to hit a chosen target (BAC or EAC). A TCPI above about 1.05 with no credible recovery plan typically signals the budget is no longer achievable.

EV can only be earned against a defined scope element. Without a stable, baseline-aligned WBS the numerator (EV) and denominator (PV) lose meaning and variance becomes noise.

LOE earns value on a time-proportional basis equal to PV, which means it never generates variance. Limit LOE to genuine support work (project management, document control) and never to deliverable scope.

The time-phased budget for all authorised scope, excluding management reserve. It is the reference for EV calculations and is changed only through formal change control.

Yes, by treating each sprint as a control account and earning value against story-point completion or feature delivery rather than activity finish. The math is identical; the unit of measure changes.

Because PV approaches BAC and EV approaches BAC by definition. This is the structural weakness of classic SPI and the reason Earned Schedule was developed.

Weighted milestones are usually best — for example IFR 25%, IFC 60%, As-Built 90%, Approval 100%. Avoid earning value on hours expended; design hours bear little relationship to engineering output.

Convert PV and EV into cash-out and cash-in profiles using payment terms, retention and lag between accrual and payment. EAC then drives the funding requirement curve seen by the CFO.

EVM measures past performance; risk drives forecast adjustments. A defensible EAC blends performance-driven projection with a quantified risk reserve so the forecast reflects both reality and exposure.

Because the same numbers flow into investor reporting, government cost reimbursement and contract claims. Weak earning rules or arbitrary % complete entries expose the organisation to restatement risk.

A projection extrapolates current performance mechanically. A forecast adds management judgement, risk exposure and committed recovery actions. Decision-makers need forecasts; analysts deliver projections.

Monthly at a minimum, in alignment with the cost report. On fast-moving projects or when major risks crystallise, refresh in-cycle and circulate a forecast amendment.

A range forecast (P10/P50/P80) produced by Monte Carlo simulation across cost, schedule and risk inputs. It communicates uncertainty rather than false-precision single numbers.

P50 is the median outcome — there is a 50% chance of overrun. Most organisations commit at P70 or P80 and hold the gap as schedule or contingency reserve.

Productivity Factor (earned hours ÷ actual hours) drives the remaining-hours forecast. A trending PF below 1.0 should be applied to all remaining hours, not just the next month.

When current performance is unrepresentative — mobilisation, ramp-down, weather seasons, or after a major sequence change. Use weighted historical performance instead of the last cycle's run-rate.

Combine EAC with payment terms, retention release and invoicing lag, then layer change-order timing and milestone payment triggers. The result is the funding curve the owner must finance.

A formal process to identify and quantify potential cost or schedule deviations before they crystallise as change orders. Trends become candidates for the next forecast update or contingency draw.

Reflect them at most-likely value with a noted probability range, then disclose them separately in the forecast narrative so management can see both committed and exposure positions.

Optimism bias, anchoring on the previous forecast, late recognition of risk crystallisation and political pressure to protect the EAC. Independent assurance and probabilistic ranges counteract drift.

Industry consensus is that float belongs to the project and is consumed on a first-come-first-served basis. Some contracts allocate float explicitly; absent that clause, the AACE 29R-03 view applies.

Total float is slack to the project finish or next constraint. Free float is slack before any successor's early start is affected. Free float is always ≤ total float.

Negative float occurs when a constraint requires the work to finish earlier than the logic allows. It is the algebra of an impossible plan and must be resolved through scope, sequence or contractual change.

Slowly — non-critical activities slip into their float without triggering alarms until they become critical. Monitor float trends weekly, not just at the data date, to catch erosion early.

A path with float within a defined threshold (commonly 10 working days) of the critical path. Near-critical paths frequently become critical mid-project, so monitor them with equal rigour.

Always specify. CPM calculates in the activity calendar — typically working days. Reporting calendar days to executives without conversion creates avoidable miscommunication.

No — a baseline with negative float is non-conforming and should be rejected. Negative float at baseline indicates the contract milestones cannot be met as planned.

Comparing the float profile across schedule updates to see whether contractor delays, owner-caused events or third-party events are eroding contingency time.

AACE RP 29R-03 defines nine forensic methods including As-Planned vs As-Built, Impacted As-Planned, Time Impact Analysis (TIA), Window Analysis, and Collapsed As-Built. Method selection depends on schedule quality, data availability and contract.

When a discrete delay event needs prospective evaluation against a contemporaneous schedule. TIA inserts a delay fragnet and measures the critical path movement attributable to the event.

A retrospective method that divides the project into time slices (windows). For each window the critical path is established, then the change in completion date is attributed to the event(s) driving the critical path.

Two independent delay events — typically one excusable and one non-excusable — affecting the critical path during the same period. Different jurisdictions allocate time and money very differently, so the contract law matters.

An excusable delay grants additional time. A compensable delay grants both additional time and cost. The contract defines which events fall in each bucket.

A UK-origin protocol widely referenced internationally for delay and disruption analysis. It promotes contemporaneous time impact analysis and structured disruption claims based on measured productivity.

A simple comparison of the baseline schedule against the as-built record to identify variance windows. Best for low-complexity projects with clean records; weak on concurrency.

A retrospective technique that removes alleged delay events from the as-built schedule to estimate the but-for completion date. Sensitive to logic inserted into the as-built network.

Forensic analysis loses credibility if the underlying records were created after the event. Daily progress, minutes, RFIs, weather logs and inspection reports must be captured in real time.

Yes — over-detailed schedules with thousands of micro-activities make the critical path volatile and conceal the dominant delay drivers. Strip the network to a defensible level of detail before analysis.

A small network of activities representing a discrete delay event, inserted into a contemporaneous schedule to measure the prospective critical path impact.

Time-related site overhead — supervision, accommodation, equipment, insurance — incurred because the project is on site longer than planned. It is the principal cost head in most EOT claims.

Loss of productivity caused by interference with the planned method of working. Disruption is distinct from delay and is typically quantified via the measured-mile or industry productivity studies.

Long enough to be defensible. Most submissions include narrative, schedule exhibits, fragnets, document references and an opinion. Quality of evidence matters more than page count.

Non-contemporaneous data, hindsight logic, ignoring concurrent events, undisclosed assumptions, and using methods inconsistent with the contract. Any of these can collapse the claim in arbitration.

A PMO defines methodology, owns delivery standards, runs portfolio reporting and assurance, and uplifts project management capability. The best PMOs measure their value in adoption and decision quality, not document volume.

Supportive (templates and coaching), Controlling (mandatory standards and assurance) and Directive (PMs report into the PMO). The right archetype depends on portfolio risk and organisational maturity.

Across governance, methodology, data quality, capability development, portfolio analytics and stakeholder satisfaction. Maturity models like P3M3 provide standardised assessment.

Portfolio CPI/SPI, on-time delivery rate, change order intensity, risk burn-down, milestone hit rate and forecast accuracy. Avoid vanity metrics like number of templates.

When a project breaches its EAC tolerance, when forecast accuracy degrades materially between cycles, when the risk reserve drops below a defined threshold, or when assurance findings are unresolved beyond their due date.

By providing portfolio visibility across agile and traditional projects, integrating sprint metrics into governance reporting, and protecting funding cadence so teams are not destabilised by annual budget cycles.

Ranking initiatives against strategic objectives, capacity, dependency and risk so capital is deployed to highest-value work. A defensible portfolio model combines a scoring rubric with a constraint engine (capacity, talent, cash).

Over-investment in process, under-investment in analytics, no executive sponsor, no consequence for non-compliance, and a focus on policing rather than enabling.

A formal decision point where the project must demonstrate readiness to proceed to the next phase — typically business case, design, execution, handover. The PMO usually convenes the gate.

A programme is a coordinated group of related projects delivering shared outcomes. A portfolio is the set of all programmes and projects competing for the same resources and capital.

Project management owns delivery and decisions. Project controls owns the information system that makes those decisions defensible — schedule, cost, risk, change, reporting and forecasting.

SPI, CPI, productivity factor, change order intensity, RFI turnaround, NCR rate, safety TRIR and forecast cost at completion. These ten numbers tell a director almost everything they need to know each month.

Through quantity surveyors who measure installed work against schedules of rates, often supported by mobile field tools that capture progress at the point of installation rather than via end-of-month aggregation.

A formal contract instrument used to add, omit or change scope. It must capture description, time impact, cost impact and basis of valuation, and is the primary mechanism for managing change in lump-sum and re-measurable contracts.

A structured assessment of subcontractor performance across schedule reliability, productivity, quality, safety and commercial conduct. It informs payment, future tender lists and improvement plans.

A time-phased plan for tendering, award, fabrication and delivery of long-lead items. Procurement delays drive a disproportionate share of construction overruns and must be visible on the critical path.

Typically by physical quantity installed against the BOQ — cubic metres of concrete, metres of pipe, tonnes of steel — converted to percent complete weighted by value.

A separate, system-based schedule that tracks loop checks, energisation, performance tests and acceptance. Mature contractors run it as a parallel network linked to construction completion of each system, not the overall project.

Measuring installed work, valuing variations, negotiating subcontractor claims and producing the monthly application for payment. The QS and the cost engineer collaborate to keep the cost report and the application aligned.

A commitment is a purchase order or contract awarded. An accrual is the value of work performed but not yet invoiced. Both are tracked to give a true forecast cost at completion.

A productivity loss technique that compares productivity in an undisrupted period of the same project against productivity in the disrupted period to isolate the disruption impact.

They are the contemporaneous record of weather, workforce, equipment, deliveries, instructions and progress. They underpin every EOT claim, disruption claim and dispute resolution submission.

A risk is a future uncertain event with potential impact on objectives. An issue is a present problem requiring management response. Risks belong on the register; issues belong on the issue log.

Quantify identified risks using probability × impact (EMV) and run a Monte Carlo simulation to derive the cost and schedule distributions. Set contingency at the chosen confidence level (typically P70 or P80).

Contingency covers known unknowns inside the risk register. Management reserve covers unknown unknowns and is held outside the PMB at executive discretion.

By scoring probability and impact on a 5x5 matrix and ranking the register by exposure. It is fast, accessible and adequate for prioritisation, but inadequate as the sole basis for contingency sizing.

A probabilistic technique that runs thousands of iterations using random samples from input distributions to produce an output distribution (cost, schedule, NPV). The output is a probability curve rather than a single number.

How quickly a risk can materialise once triggered. A high-velocity risk leaves little time to respond and may demand pre-positioned mitigation even at lower probability.

Avoid (eliminate the cause), Transfer (insure or contract out), Mitigate (reduce probability or impact) and Accept (acknowledge and hold reserve). Opportunities mirror these as Exploit, Share, Enhance and Accept.

The named individual accountable for monitoring a specific risk and executing the agreed response. Without explicit ownership, register entries become administrative noise.

Because conditions change — design matures, suppliers fail, regulations shift. Static registers lose credibility within months; effective registers are re-scored monthly with named change drivers.

A Monte Carlo simulation that randomises both activity durations and cost inputs, capturing the correlation between them. It exposes the true funding curve a probabilistic schedule will require.

Inherent risk is exposure before any controls are applied. Residual risk is what remains after controls. Boards typically govern against residual risk thresholds.

Opportunities are positive risks and should be registered alongside threats, with explicit Exploit/Share/Enhance responses. Mature portfolios actively pursue opportunity capture rather than treating it as upside luck.

A visual that tracks the closure of risks against time, useful for showing whether the register is being actively managed or simply accumulating entries.

Too many risks, too little prioritisation, no link to financial exposure and no clear ask. Board-grade risk reporting summarises top exposures in financial terms with a recommended decision.

Aligning project-level risks with the enterprise risk register so escalation paths, reserves and reporting are consistent across the organisation rather than fragmented.

Entitlement under the contract. Without a contractual basis the merits do not matter. Identify the clause, the trigger event, the notice obligation and the relief mechanism before assembling quantum.

Contemporaneous schedules, daily reports, weather logs, RFI logs, minutes, change orders, photographic evidence, productivity records and a clean baseline. Records assembled after the fact rarely survive scrutiny.

A defined period (often 28 days under FIDIC) within which the contractor must notify the engineer of the claim event. Missing notice can extinguish entitlement entirely under strict contracts.

Through head-by-head substantiation of prolongation costs — site overheads, supervision, equipment, insurance — supported by the cost system and apportioned to the delay period.

A claim that pleads cumulative impact without isolating the contribution of each event. Courts and adjudicators routinely reject global claims when discrete analysis was possible.

A claim is an assertion of entitlement under the contract. A dispute arises when the claim is rejected and the parties cannot agree. Disputes move into mediation, adjudication or arbitration.

Delay claims address completion-date impact; disruption claims address lost productivity even when completion is unchanged. Both require contemporaneous records and recognised quantification methods.

A standing panel appointed at contract inception (typical in FIDIC and infrastructure contracts) to resolve disputes quickly during execution and avoid post-completion arbitration.