Manhour Budget and Productivity Tracking
Last reviewed 9 September 20262,231 words10 min read
Earned hours over actual hours, weekly, by trade — the number that forecasts the schedule before the schedule does
🟩 In one line: Track every trade weekly as earned manhours (installed quantity × budget rate) against actual manhours (timesheets by cost code); the ratio is the productivity factor, it drives your Remaining Durations and your manhour forecast to complete, and it is the only contemporaneous record a measured-mile claim can be built on.
👤 Who this is for: Mid-level planners running progress; senior planners setting up cost codes with the commercial team at mobilisation. Prerequisites: Crew-and-manhour-formulas, Quantity-tracking-and-physical-progress, Resource-loading-and-histograms.
First, let's be honest about why this page exists
The duration of every production activity came from a unit rate. If the site is achieving 0.42 mh/m² of blockwork against a budget of 0.35, every remaining blockwork activity is 20 % longer than the schedule says, and no amount of statusing will reveal it until the dates slip. Productivity tracking is the leading indicator; the schedule update is the lagging one.
It is also the record. Two years later, when the disruption claim says congestion cost 25 % on Levels 8–15, the only evidence that survives cross-examination is weekly earned-versus-actual hours by trade, by area, kept from week one. Nobody can reconstruct it afterwards.
🟨 The standard — what "good" looks like
| Source | What it says (paraphrased) | Use it for |
|---|---|---|
| AACE RP 25R-03 Estimating Lost Labor Productivity in Construction Claims | Measured mile preferred; requires contemporaneous productivity data by work type and period; lists loss causes (check against your copy) | Why the record must be weekly and by trade |
| SCL Protocol 2nd ed., Guidance Part B (disruption), Core Principle on records | Disruption proven by comparing achieved productivity in impacted and un-impacted periods; records kept contemporaneously (check against your copy) | The evidential standard |
| FIDIC 1999 Cl 6.10 / 2017 Cl 6.10 | Contractor's records of personnel and equipment on site, by category, submitted monthly (check against your copy) | Contractual headcount record |
| FIDIC 1999 Cl 4.21 / 2017 Cl 4.20 | Progress report content includes personnel and equipment and comparison of planned and actual progress | Where the productivity summary goes |
| NEC3 / NEC4 Cl 31.2 (resources on programme); Option C/E Defined Cost records | Resources planned per operation; cost records for Defined Cost | Manhour records feed both |
| AACE RP 38R-06 | Productivity assumptions and unit rates documented in the SBM | Budget rates traceable |
| CII benchmarking (Construction Industry Institute) | Productivity metrics defined as workhours per unit installed, by discipline | Metric definition, neutral context only |
| Hub convention | PF = earned ÷ actual (1.00 = on budget; > 1 better); trend on last 4–6 weeks; RD = remaining qty ÷ achieved rate; quantity sheet is the source; one workbook, tab per trade | Your defaults |
🟢 Rule: budget rate × installed quantity = earned hours; timesheets by cost code = actual hours; the weekly ratio by trade is the productivity factor, and every Remaining Duration in the next update is derived from it.
How it actually works
1. Set up the structure at mobilisation, with the commercial team.
| Element | Decision |
|---|---|
| Cost codes | One per trade per area group, aligned to WBS Level 3 and the SBM duration-basis tables (e.g. STR-L01-L10-REBAR); 60–150 codes on a tower |
| Budget hours | Per cost code: quantity × SBM unit rate; the sum equals the labour BAC loaded in P6 |
| Timesheet | Daily, by worker, by cost code, hours actually worked (not paid — overtime premium is money, not hours); subcontractor labour recorded the same way under the subcontract appendix |
| Quantity sheet | Installed and IR'd quantities per activity weekly; the earned side |
| Ownership | Timesheets: site admin / cost engineer; quantities: planner with QS; workbook: planner |
2. The weekly calculation, per trade and per area.
| Column | Content |
|---|---|
| Budget qty / budget mh / budget rate | From SBM |
| Installed qty to date | Quantity sheet |
| Earned mh to date | Installed × budget rate |
| Actual mh to date | Timesheets |
| PF (cumulative) | Earned ÷ actual |
| PF (period, last 4–6 weeks) | Period earned ÷ period actual — the trend that matters |
| Achieved rate | Actual ÷ installed (mh per unit) |
| Remaining qty | Budget − installed |
| Forecast mh to complete | Remaining qty × achieved rate (period) |
| Forecast total mh | Actual + forecast to complete |
| Variance mh | Forecast total − budget |
3. Push it into the schedule. At each DD, for every in-progress and not-started activity in a trade: RD = remaining qty ÷ (crew × hours/day ÷ achieved rate) — the same formula as Forecasting-honestly, with the achieved rate from this workbook. Remaining Units = remaining qty × achieved rate. Both entered separately; Notebook Forecast Basis names the workbook tab and the PF used. Not-started activities in a trade with sustained PF < 0.90 get their Original Duration left as baseline and their RD re-forecast; the difference is reported as duration growth with cause tag.
4. Put actual hours in P6. Monthly, Actual Units per resource assignment from the workbook (spreadsheet import, never retyped), so CPI in hours and the histogram's actual bars are real. Remaining Units updated at the same time; At Completion Units = actual + remaining and the variance against Budgeted Units is visible per activity.
5. Read it for causes, not just numbers. A PF drop is diagnosed the same week:
| Pattern | Likely cause | Check |
|---|---|---|
| PF falls in one area, others steady | Access, congestion, rework, information | Look-ahead constraints list, RFIs, IR rejections for that area |
| PF falls across a trade everywhere | Crew change, overtime fatigue, supervision | Regime table on Overtime-and-shift-arithmetic; gate log |
| Actual hours flat, earned hours flat | Standing time | Daily diary; notice under Cl 20 / Cl 61 if Employer-caused |
| Earned rising, actual rising faster | Overmanning | Histogram by zone; second-crew factor |
| PF fine, RD still slipping | Crew size below plan, not productivity | Compare actual headcount to Max Units |
6. Keep the un-impacted period clean. A measured mile needs a stretch of the same work, same trade, same conditions, without the disruption. Tag each week per area in the workbook with an Impact flag and a cause code (design change, access, weather, Employer instruction, none). Without the tag the analyst cannot find the mile later.
📜 Records that survive
- Weekly, by trade, by area, from week one. Reconstructed productivity is rejected under RP 25R-03 and the Protocol alike; a workbook with a file date each week is the evidence.
- Hours worked, not hours paid. Overtime premium inflates cost, not hours. A claim in hours converted at real rates is auditable; one in paid-hour equivalents is not.
- Subcontractor hours under the appendix. If the subcontractor does not report actual hours by cost code, the main contractor cannot support a disruption claim on their behalf and cannot resist one from them. It goes in the subcontract appendix (Subcontractor-and-interface-integration).
- Notice. A PF collapse caused by the Employer is a claim event the week it starts (FIDIC 28 days; NEC 8 weeks). The workbook dates the start.
🟥 Where people go wrong
- Hours by headcount only. Gate log × 10 hours, no trade, no cost code. It tells you how many people were on site and nothing about what they did. Timesheets by cost code from day one; the gate log is the cross-check.
- PF measured on paid hours. A 12-hour day paid at 14 hours looks 17 % worse than it is. Hours worked.
- Budget rates from the estimate, not the SBM. The estimator's rate included allowances the SBM's duration rate did not, and the PF says 0.85 for a trade that is on programme. One rate per trade, in the SBM, used for duration and for earned hours.
- Cumulative PF only. Twelve good months hide a bad quarter. Period PF on the last 4–6 weeks drives the forecast; cumulative is the trend line.
- Earned hours from percent. "Blockwork 60 %" × budget hours. Percent came from the quantity sheet; use the quantity. Percent-derived earned hours are circular.
- Productivity workbook not linked to Remaining Durations. The workbook says 0.78 and the schedule's RDs still assume 1.00. The whole point is the RD refresh at each DD, Notebook-referenced.
- No impact tags. Two years on, nobody can say which weeks were disrupted, so there is no measured mile. Tag every week per area with a cause code, including "none".
⚖️ When you're challenged
"We have 1,800 people on site against 1,500 planned — why are we behind?" Because the productivity factor on the finishing trades is 0.74 — for every budgeted hour of work we're spending 1.35. Eighteen hundred people at 0.74 produce less than 1,500 at 1.00. The workbook shows it started on Levels 8–15 in the week the ceiling design change landed.
"Where does your remaining duration come from? The subcontractor says three weeks." From their achieved rate: 4,200 m² remaining at 0.46 mh/m² is 1,930 hours; their crew of 24 at 10 hours is 240 a day; that's 8.1 days — call it 9 — per zone, three zones, 27 working days. Their three weeks assumes the budget rate they haven't hit since April.
"Just show me one productivity number for the project." Cumulative PF is 0.91 and period PF is 0.83. The first says where we are; the second says where we're going, and it's the one driving the forecast. Both are in the report by trade.
"Can we claim disruption for the congestion on the podium?" The workbook shows blockwork PF of 1.02 on Levels 2–5 before the Employer's re-sequencing instruction and 0.71 on Levels 6–9 after it, same trade, same subcontractor, same crew. That's the measured-mile comparison the analyst will need; the notice went in the week PF dropped.
📄 Related pages
- Crew and Manhour Formulas — the budget rate the earned hours use
- Quantity tracking and physical progress — the installed quantities
- Forecasting honestly — RD from achieved rate, entered with Remaining Units
- Productivity Adjustment Factors — planned factors vs the achieved PF
- Overtime and Shift Arithmetic — when the PF drop is the regime, not the site
- Measured mile and disruption — what the workbook becomes in a claim
- Earned Value Basics: SPI and CPI — the same earned hours feed EV
- Resource Loading and Histograms — the budgeted units the actuals are measured against
✏️ Worked example — Dubai 22-storey tower, blockwork, week ending 22-Nov
Budget rate 0.35 mh/m² (SBM Table 5-3). Crew 24 on 6-day 10 h.
| Week | Area | Installed m² | Earned mh | Actual mh | Period PF | Cum PF | Impact tag |
|---|---|---|---|---|---|---|---|
| 40 | L05–L06 | 1,620 | 567 | 560 | 1.01 | 1.03 | None |
| 41 | L06–L07 | 1,580 | 553 | 570 | 0.97 | 1.02 | None |
| 42 | L07–L08 | 1,540 | 539 | 590 | 0.91 | 1.00 | None |
| 43 | L08–L09 | 1,180 | 413 | 610 | 0.68 | 0.95 | Design change — ceiling void revision, L8–L15 (VO-014 pending) |
| 44 | L09 | 1,090 | 382 | 620 | 0.62 | 0.90 | Design change; access — hoist down 2 days |
| 45 | L09–L10 | 1,210 | 424 | 615 | 0.69 | 0.87 | Design change |
| 46 | L10 | 1,260 | 441 | 600 | 0.74 | 0.85 | Design change |
Trend (weeks 43–46): period PF 0.68; achieved rate 0.51 mh/m². Remaining blockwork L10–L15: 8,900 m².
| Item | Baseline basis | Re-forecast |
|---|---|---|
| Remaining mh | 8,900 × 0.35 = 3,115 | 8,900 × 0.51 = 4,539 |
| Daily output (24 × 10 = 240 mh) | 13.0 d | 18.9 → 19 d |
| Variance | — | +6 working days on the L10–L15 chain |
Actions: RDs on BLK-L10 to BLK-L15 re-forecast at 0.51 mh/m² (Notebook: "Forecast Basis — productivity workbook tab BLK, PF 0.68 wks 43–46"); Remaining Units set to 4,539 across the six activities by zone; longest-path effect +6 days, reported with cause tag duration growth — design change VO-014; notice N-021 issued in week 43 references the PF drop; weeks 40–42 flagged as the un-impacted comparison period. Monthly, actual units 4,165 mh imported to P6 against the six blockwork assignments.
📖 References
- AACE International RP 25R-03, Estimating Lost Labor Productivity in Construction Claims (check the edition in your contract)
- SCL Delay and Disruption Protocol, 2nd ed. (2017), Guidance Part B — disruption; Core Principles on records (check against your copy)
- FIDIC Conditions of Contract for Construction 1999, Cl 4.21, 6.10; 2017, Cl 4.20, 6.10 (check the edition in your contract)
- NEC3 / NEC4 ECC, Cl 31.2; Options C and E, Defined Cost records (check against your copy)
- AACE International RP 38R-06, Documenting the Schedule Basis
- Oracle Primavera P6 Professional User Guide — Resource Assignments; Actual Units; spreadsheet import (check against your P6 version)
Illustrative ranges, for sanity-checking only. Not a substitute for your own as-built data, and not for use as evidence in a contractual claim.
From the field
Experience from working planners. Unreviewed — read it as experience, not guidance.
Add what you know about manhour budget and productivity tracking. What worked, what the consultant pushed back on, what you would do differently next time. A paragraph is plenty.
Contributors get their name and one link on the site — your own templates, course or consultancy. We take nothing and hold nothing.
Add your experience