Prepare a Construction Final Account Workbook (UK QS Guide)

The construction final account is the agreed conclusion of the contract sum and the figure the employer will pay the contractor. In operational terms it is a single arithmetic statement:

Final Account = Contract Sum + Additions − Omissions − Retention Release Net + Loss and Expense + Fluctuations

Additions consolidates three streams: the variations register at agreed value, the settled dayworks total, and the actual expended amount on every provisional sum and prime-cost sum. Omissions captures de-scoped work, unused provisional sums formally instructed away, and any negative variations. The result, reconciled against cumulative interim payments certified through the project, is the final balance owing to the contractor — or, on omissions-heavy projects, owed back to the employer.

To prepare a construction final account workbook that holds up in agreement and in negotiation, build it as a summary sheet plus one section per adjustment stream: variations, dayworks, provisional sums, prime-cost sums, fluctuations, loss-and-expense, retention release. Each section totals to a single line on the summary sheet. Every row in every section ties to a source document — a contract administrator's instruction, a signed dayworks sheet, a supplier or subcontract invoice, an interim certificate, a retention ledger entry. The summary sheet reads as a single page that any reviewer on the other side can follow without a walk-through.

The opening figure on that summary sheet is the contract sum itself, taken from the contract sum analysis or the priced bill of quantities. Where the priced BOQ exists only as a PDF — as it often does on the employer's-agent side or where the original tender file has been lost — the sensible first move is to convert the priced NRM2 bill of quantities into Excel so the contract-sum opening line carries the same row structure the rest of the workbook will be built against.

What follows walks each adjustment stream in the order the workbook builds: variations consolidated from the live register, dayworks settled at defensible rates, provisional and prime-cost sums replaced with actual expenditure, fluctuations and loss-and-expense recorded as agreed lines, retention released into the final balance, and the cumulative-paid reconciliation that produces the closing figure. The discipline is consistent throughout: every line ties to a document, every section total reconciles upward, and the workbook can be opened by the other side without explanation.

Consolidating the Variations Stream from the Live Register

The variations register is the upstream feed; the final-account variations section consolidates it into a single line on the summary sheet. It does not replicate the register, and it is not the place to value disputed entries for the first time. Every row that lands here should already be at agreed value. Anything still in dispute, awaiting CA valuation, or sitting in the queue as an instructed-but-unpriced variation needs to be resolved before the workbook can close — or recorded explicitly as a contingent adjustment with a stated assumption, so the other side can see what is in and what is parked.

The mechanism by which a variation arrives at agreed value depends on the contract. There is a real difference between the JCT and NEC routes, and a final-account workbook that conflates them tends to leak credibility on the variations section first. RICS describes the difference plainly: RICS Final Account Procedures (1st edition) practice information notes that the NEC suite of contracts does not refer to final accounts because it is assumed that the final account is adjusted as the project proceeds on the basis that the compensation event procedure is followed; the JCT defines the final account process under clause 4.12 of the Design and Build Contract or clause 4.5 of the Standard Form.

Under JCT, variations are valued retrospectively by the contract administrator under the clause 5 family in the Standard Building Contract — or its equivalent valuation provisions in the Design and Build, Intermediate, and Minor Works forms. The Schedule 2 valuation rules sit underneath: measure at bill rates where the varied work is of similar character executed under similar conditions, fair valuation where it is not, and dayworks where measurement is impractical. In a well-run job, most of the register has already been valued and signed off through the project; at final account, the QS is confirming the agreed values and resolving the residual queue.

Under NEC, variations sit inside the compensation event procedure — clause 60 family — and are agreed prospectively with the Project Manager as the work proceeds. Each compensation event is notified, quoted, assessed, and implemented before the work is executed where the contract is being run as it is intended to be. The register entries that flow into the final-account workbook should already carry quotation references and implementation dates, with no retrospective valuation backlog. Where the project has accumulated unimplemented compensation events — a common pathology — those need to be cleared, not parked into the workbook unresolved.

From the agreed-value register, produce the additions / omissions summary by instruction reference. Each variation instruction is a row, with the agreed value split into an additions column or an omissions column depending on whether it adds work or removes it. The columns total separately, and the net of the two is the single line that flows up to the summary sheet. Keeping additions and omissions in their own columns matters at negotiation: the gross figures are usually what is interrogated, not the net.

De-scope omissions deserve their own attention. Where significant scope has been pulled out — sections of the works deleted, packages descoped, instructed reductions to specification — the omissions column can be substantial enough to drive the final account below the contract sum. The workbook must be capable of carrying a negative net variations figure with the same source-document discipline as a positive one. Where the project has driven enough omissions that the contractor's commercial position is exposed, the variations section is also where the QS team should cross-reference internally to the project-end position; commercial managers preparing for final-account negotiation typically want to build the project-end cost value reconciliation spreadsheet alongside the final account so the negotiating position is grounded in the contractor's actual cost outturn, not just in the contract entitlement.

A small number of high-value variations almost always drive the consolidated figure. It is worth tagging the cost-significant items in the workbook — the top five or top ten by absolute value, including significant omissions — and cross-referencing each to its underlying CA instruction, valuation paper, and supporting measurement or quotation. When the other side opens the workbook, they will ask about those lines first. The supporting evidence sitting one click away from each row is what makes the variations section defensible.

Settling the Dayworks Stream at Defensible Rates

Dayworks is where the workbook earns or loses its credibility. The variations section consolidates work that has already been valued; the dayworks section is the QS doing the assessment, often against a stack of signed sheets where the contractor's claimed labour, plant, and materials need to be tested against the contracted rates and the actual evidence of what was on site.

The controlling reference is the agreed dayworks rate sheet at contract level. That document — built into the contract or appended via the priced preliminaries — sets out four things: the labour-rate schedule (typically gangs by trade and grade, with all-in rates that include base pay, statutory on-costs, and small tools), the plant-rate schedule or hire-rate basis (with operator inclusion or exclusion stated explicitly so a hired-with-driver item is not double-counted against an operator labour line), materials at net invoice cost, and the on-cost percentages that apply to each. The on-cost convention varies but is commonly a percentage on labour for supervision, site overheads, and head-office overheads-and-profit, and a percentage on materials and plant for handling, wastage, and overheads-and-profit. The contracted rates and percentages are the only rates and percentages the workbook should use; any sheet claimed at a higher level is assessed back to the contract level as part of the settlement.

Through the project, the QS or site engineer should have built up a signed-sheet trail. A dayworks sheet signed by the CA, employer's agent, or site engineer at the time the work is being carried out — recording labour hours by gang and grade, plant hours by item, and materials with delivery references and supplier-invoice numbers — is the strongest evidence the contractor has. Sheets signed retrospectively, sometimes weeks after the operation, are weaker; partial sheets where a signature confirms attendance but not the claimed quantum are weaker still. The settlement workbook needs to track each sheet's signature status, because the assessment criteria differ depending on what evidence the sheet carries.

For each sheet that goes into assessment, the criteria are evidential rather than negotiated:

• Labour hours validated against site presence. Cross-check the claimed gang against the signing-in records, the attendance register, and the daily allocation sheets. A gang of four claimed for a full shift on a dayworks operation needs four operatives signed in for the shift and allocated to that operation by the supervisor. Hours that cannot be evidenced as present and engaged on the operation in question are reduced or rejected. The all-in labour rate from the contract schedule is then applied to the validated hours.
• Plant hours validated against hire records or owned-plant logs. A claimed item must be both present on site and engaged on the dayworks operation for the claimed period. For hired plant, evidence is the hire invoice with start and off-hire dates that bracket the dayworks period; for owned plant, evidence is the plant utilisation log showing the item was deployed on the operation. Standing time on site is not the same as engaged time on the operation.
• Materials at net invoice cost. The supplier invoice is the evidence. The net amount before any settlement discount becomes the assessed value, multiplied by the contracted on-cost percentage for materials. Trade discounts that the contractor has actually received reduce the net cost; a quoted list price that bears no relationship to the invoice is not a defensible basis.

Unagreed sheets — the sheets the contractor has submitted but the CA has not signed at the time of the work — are the area where the assessment most often slips into argument. The discipline is the same: assess at contract dayworks rates, against the same evidence streams, with claimed lines accepted only where the evidence supports them. Each rejected or reduced line needs a clear narrative in the workbook — what was claimed, what was assessed, why — because that narrative is what the QS will be asked to defend in the negotiation. A workbook that says "rejected: insufficient evidence" without naming what evidence was sought and not produced does not survive cross-examination.

The output of all this work is a single line on the summary sheet: the consolidated dayworks total, settled at contract rates against the available evidence. The detail — per-sheet assessment, per-line evidence, per-line rate calculation, signed-sheet status, supporting documents indexed — lives in the dayworks section of the workbook. When the other side asks how the line was built up, the section opens to the per-sheet assessment, and each line opens to its evidence.

The aggregation problem itself is real. On a job of any meaningful length, the dayworks stream can run to hundreds of signed sheets, each one a PDF, plus the supplier invoices for materials, the hire records for plant, and the daily allocation sheets that validate labour. Pulling that volume into structured rows in the workbook — sheet number, date, operation, gang composition, claimed hours, claimed plant, claimed materials, supporting invoice references — is the data-handling step that often holds the section up. Some QS teams type each sheet by hand; others use the product behind this site to extract structured data from construction PDFs into the workbook, processing batches of signed sheets and the supporting supplier invoices through a single prompted extraction so the dayworks section starts populated and the QS time goes into assessment, not into transcription. Whichever route the team takes, the assessment that follows is the same — the rate sheet, the evidence streams, the line-by-line narrative — and what the workbook ultimately defends is the assessment, not the data-entry method.

Settling Provisional Sums and Prime-Cost Sums Against Actual Expenditure

Provisional sums and prime-cost sums are placeholders. The contract sum already includes them at their tendered figure; the final-account workbook replaces those included figures with what was actually expended.

NRM2 distinguishes two types of provisional sum, and the distinction matters for how the contractor is entitled to be paid. A defined provisional sum is for work that is not fully designed at tender stage but for which sufficient information is provided — nature and construction, statement of how and where the work fits, quantity, and any specific limitations — for the contractor to have made allowance in their programme, planning, and pricing of preliminaries. An undefined provisional sum is included where that information is not available; the contractor is entitled to the full impact on programme and preliminaries when the sum is expended via instruction. The QS preparing the workbook needs to know which type each provisional sum was, because the entitlement that flows from settlement is different.

The settlement mechanism itself is the same for both types. The contract sum carries the included figure; the workbook strips that included figure out and replaces it with the actual expended figure, sourced from the contract instruction expending the sum and the supplier or subcontract evidence behind it. Where the work was carried out by the main contractor with their own labour and plant, the expended figure builds up from the priced instruction and the supporting cost records. Where the work was sublet, the expended figure is the agreed value of the subcontract package against the original provisional. Either way, the row records the included figure, the expended figure, the net adjustment, and the source.

Unused provisional sums require explicit action. They do not simply lapse off the contract sum because they were not spent. The contract administrator must instruct the omission, and the workbook records the omission as a row with the included figure in the omissions column and the source instruction in evidence. A workbook that just leaves an unused provisional out of the additions stream without instructing the omission is not balanced — the contract sum still includes the figure.

Prime-cost sums work the same way mechanically but with nominated parties on the other side. A PC sum sits in the contract sum for a nominated subcontract or nominated supplier element. Settlement replaces the contract-sum-included PC with the actual nominated figure — the agreed final account on the nominated subcontract, or the supplier invoice for nominated materials. The contractor's profit and attendance percentages set out in the contract are then applied to the actual PC figure, not the included PC, because the basis of the percentages is the actual cost flowing through.

The workbook treatment is consistent with the rest of the spreadsheet. A row per provisional or PC reference, with columns for the included figure, the expended figure, the net adjustment, the source instruction reference, and the supporting evidence — invoice number, subcontract reference, or nominated-supplier order. The provisional-sums section totals to one line on the summary sheet; the PC-sums section totals to another.

The supporting evidence is invoice-heavy. Every expended figure is built up from supplier invoices, subcontract applications, or nominated-supplier orders, and the cost ledger needs to carry those invoices coded to the correct provisional or PC reference so the workbook total reconciles to the project cost report. Teams that have been disciplined enough to code construction supplier invoices to job and cost code through the project will find the provisional and PC reconciliation builds straight out of the cost ledger. Teams that have coded everything to a generic project code will spend the time at final account doing the recoding the cost-control discipline should have done in real time.

Applying Fluctuations and Loss-and-Expense

Fluctuations and loss-and-expense are the two streams the workbook records as agreed lines rather than as detailed build-ups. Both can be substantial in money, but on most projects the build-up sits outside the workbook itself — fluctuations in an indexation calculation, loss-and-expense in a separately negotiated submission — and what the final account carries is the agreed figure with a clear pointer back to the underlying paper.

Fluctuations are clause-dependent and frequently inactive. Most lump-sum contracts in the UK market are let on a fixed-price basis with the fluctuations provisions disapplied; on those projects the workbook line is nil. Where the contract permits formula-based adjustment — typically the NEDO Price Adjustment Formulae or the equivalent index-linked mechanism specified in the contract particulars — the workbook records the calculated adjustment by valuation period, with the indices used, the base date, and the net values supporting each period's figure. Where the contract permits full fluctuations, recovering actual changes in the cost of labour, materials, and statutory items, the line is supported by the full build-up: each evidenced cost change tied back to the underlying source — a wage-rate award, a manufacturer's price-change notification, a statutory uplift — and aggregated across the period claimed.

The cost-reimbursable and target-cost variants under NEC are a different animal. Under Option C and Option E, the final account is not built up from variations to a lump sum — it is built up from defined cost minus disallowed cost, with the pain/gain share calculated against the target on Option C. Mention the route, refer to the schedules of cost components and the disallowed-cost rules in the contract for the build-up basis, and proceed; a final-account workbook for an Option C project is structured around the open-book cost record rather than the contract-sum-plus-adjustments shape this article walks. The principle that every line ties to a source document still applies; the streams just sit in different places.

Loss-and-expense lands in the workbook as agreed sums, sourced from a separately negotiated process. The build-up — the extension-of-time award supporting the prolongation period, the prolongation cost itself (additional preliminaries, site management, supervision through the extension), the head-office overhead claim where the contract provides a route for it, finance charges where claimable — sits in the loss-and-expense submission and the supporting bundle. The workbook records the agreed figure as a single line, with a reference back to the L&E submission file and the agreed correspondence that closed the negotiation.

This is worth being explicit about: the workbook is not the place to negotiate loss-and-expense. By the time the figure lands on the summary sheet it is at agreed value, with the negotiation already concluded. If the L&E position is unresolved at final-account stage, the workbook either parks the line as a contingent adjustment with a stated assumption or holds the final account open until the L&E is settled. Trying to settle L&E inside the workbook itself, after the rest of the streams have been consolidated, conflates two negotiations that need to run on their own terms.

Releasing Retention into the Final Account

Retention is held against each interim valuation through the project as a percentage of certified gross value, typically 3 or 5 percent depending on the contract. It releases in two tranches: half at practical completion, the second half at the end of the defects liability period or on issue of the certificate of making good defects, whichever the contract specifies as the trigger. The final-account workbook records the release as a single line, with the underlying ledger sitting in the retention tracker.

By the time the workbook is being assembled, practical completion has been certified and the first-half release has flowed through interim certification — it sits inside the cumulative-paid figure on the reconciliation row. The final account triggers the second-half release, timed against the certificate of making good defects per the contract. The workbook line is that second half net of any retention adjustments arising from defects that were not made good.

The detailed retention ledger lives in the retention tracker, not in the final-account workbook itself. That ledger carries the per-valuation history: each interim certificate's certified gross, the retention deducted at the contractual percentage, the running retained balance, and the release events as they occur. A retention tracker that has been kept current through the project produces the final-account line straight from the closing balance; one that has not been kept current is where the QS team finds itself reconstructing the retention history from the certificate stack at final-account stage. Either way, the figure that lands on the summary sheet is the net amount being released into the final balance, with the ledger reference noted in support.

Retention bonds and trust-account variants change the cash flow but not the workbook structure: the line records the release event with the bond instruction or trust drawdown referenced in support. Sectional completion and partial possession push the same point — the retention tracker carries the per-section release history, and the workbook just rolls those positions up into the single final-account line.

Reconciling Cumulative Payments and Closing the Final Balance

With every adjustment stream consolidated, the workbook closes with the arithmetic that turns the adjusted contract sum into the final balance owing or owed. The reconciliation is mechanical: sum the certified amounts paid to date across every interim certificate (or pay-less notice as adjusted), deduct from the adjusted contract sum on the summary sheet, and the remainder is the final balance owing to the contractor. On omissions-heavy or de-scoped projects the remainder can be negative, in which case the balance is owed back to the employer.

The negative-balance case is worth treating with the same calm as a positive one. Final accounts on heavily de-scoped projects — significant omissions, unused provisional sums omitted by instruction, descoped subcontract packages — can produce a final balance below the cumulative paid, requiring the contractor to refund. Commercial conversations on those projects tend to be uncomfortable, but the workbook's job is the same: every adjustment line traceable to source, every section total reconciling upward, the cumulative-paid figure tying to the certificates that were issued. A defensible negative balance is a defensible negative balance.

The reconciliation row on the summary sheet does the closing arithmetic visibly. List every interim certificate by valuation period with its certified gross, the retention deducted, and the net certified figure. Sum the net certified figures down to the cumulative-paid total. Deduct the cumulative-paid total from the final-account adjusted contract sum and the result is the final balance line. Where the final-balance figure is being agreed against an employer's-side QS, that side's reconciliation should produce the same cumulative-paid figure to the penny — the certificates are common reference, and any difference between the two reconciliations is a counting error that needs to be tracked down before the negotiation can move on. The QS team that consolidates the cumulative-paid history correctly typically does so by working from the same monthly position the team used to assess monthly subcontractor applications for payment on the way through the project, which keeps the reconciliation rows consistent with how the project was administered in real time.

A point on the underlying paper: the certified amounts feeding the cumulative-paid figure are the AfP-driven payment certificates, not the contractor's VAT invoices. The two documents serve different purposes in the JCT and NEC payment cycle, and a reconciliation row that conflates the two will not balance. The mechanics are covered in detail elsewhere; for the purposes of the final-account workbook, the rule is simply that the reconciliation sums certified payment amounts, with the JCT and NEC application-for-payment versus VAT invoice mechanics sitting underneath as the reason the certified figure is the right figure to reconcile against.

The final-certificate timing sits one step beyond the workbook. JCT specifies the final-certificate trigger and timing in the contract particulars; NEC works to its own programme via the Project Manager. The workbook supports the certification — it is the document the certifier relies on when issuing the final certificate — but it does not replace it. What the workbook delivers is the agreed final-account position; the final certificate is the contractual instrument that crystallises payment.

What protects the workbook through the agreement and into negotiation is the discipline that has run through every section. Every line on every adjustment-stream section is traceable to a source document. Every section total ties to the summary sheet without edit. The cumulative-paid figure ties to the certificates the QS or CA issued through the project. The reconciliation arithmetic is visible on the summary sheet itself. A workbook that the other side can open, walk through without explanation, and trace from any line back to its supporting paper is the workbook that holds up under scrutiny — and it is the workbook the QS who built it can defend without having to remember what was done where.