A new study published in the journal Architectural Research on 19 October 2026 examines that proposition in detail, tracing three claim bundles from the coal-handling works of a thermal power plant in the Republic of Korea. The research, conducted by Byungyun Bae, shows that a monetary total presented in a claim is only as strong as the chain of documents behind it, and that the chain frequently breaks at predictable points.
The study addresses a common frustration in engineering, procurement, and construction projects. Claims are usually presented as single monetary figures, yet the records needed to interpret their scope, valuation, and time consequences are scattered across different document types: instructions describe the scope, quantity sheets measure the work, estimates establish the cost, correspondence records the parties’ competing positions, and schedule updates establish whether the completion date actually changed. A monetary summary alone, the author argues, cannot perform all of these functions.
To test how these records interact, the study analyzed three contrasting bundles from a single coal-handling package. Bundle A contains an additional-cost estimate and quantity take-off for an added sump-pit structure. Bundle B holds a paving-quantity summary, a detailed estimate, and a change-cost summary. Bundle C combines a stoppage cost request, supporting cost sheets, and correspondence between the parties. The author applied a documentary evidence-chain method that separates the initiating event, the version of the work and quantity, the pricing basis, the parties’ positions, and the schedule evidence into distinct analytical objects.
The method is deliberately cautious. For each bundle, an extraction matrix records five links: the stated trigger and date, the scope or quantity version used as a baseline, the arithmetic leading from work items to the requested amount, the status of each statement as a request, agreement, objection, or undetermined issue, and the contemporaneous program and work records needed for a time assessment. Each link is labeled present, contested, or unestablished, and these labels are explicitly evidentiary rather than determinations of liability. Cost figures are expressed as percentages of a specified subtotal within each bundle rather than as commercial amounts, in part because the original documents contain personal and commercial details.
Bundle A, the sump-pit addition, produced the most internally consistent picture. In the submitted estimate, materials account for 56.70 percent of the net construction cost, direct labor for 32.52 percent, and expenses for 10.78 percent, while the general-administration component is stated at 5.70 percent of that net cost and value-added tax at 10.00 percent of the supply amount. The underlying figures reconcile within the estimate. Relative to the original estimate-line quantities, concrete increases by 112.42 percent, formwork by 27.99 percent, and reinforcing steel by 237.48 percent. The quantity and estimate sheets support a priced request traceable within the accessible bundle, but the authorizing instruction and any final agreement sit outside the displayed cost sheets, leaving contractual entitlement unassessed.
Bundle B reveals a subtler problem: a presentation-basis mismatch. The documents describe a first change as completed and a further amount as requested, representing 40.37 percent and 59.63 percent respectively of the cumulative figure before value-added tax. Within the further request, a changed contracted item represents 11.64 percent, while previously unreflected or additionally measured geotextile and aggregate items represent 88.36 percent. Relative to the original quantities in the detailed estimate, slag placement, a friction-type high-density polyethylene sheet, and non-woven geotextile each increase by 11.05 percent, with the estimate starting at 7,844 cubic meters for slag and 52,297 square meters for each sheet item. However, a separate quantity summary displays starting values one unit higher while retaining the printed increments, so direct subtraction yields an increment one unit smaller in each row. Whether display rounding, a different measurement basis, or another document version explains the difference cannot be determined from the displayed records alone.
Bundle C concerns the most contested territory: time. The cumulative cost summary before value-added tax separates a first change, described as completed, at 15.22 percent from the current stoppage-related request at 84.78 percent. Within that current request, additional labor-related allowances account for 31.74 percent, equipment standby for 23.65 percent, and waiting or rework for 44.61 percent, and these percentages reconcile within the requested subtotal. The claim summary asserts a six-day impact, but the correspondence considers one day of labor allowance while contesting the additional days and the equipment and indirect-cost categories pending detailed proof. The correspondence supports the date of an interruption and shows that different causes were proposed for subsequent nonworking periods, yet the strongest supported conclusion is that the parties framed the same interruption through different duration and cost boundaries.
The stoppage analysis illustrates why delay claims demand more than cost sheets. The study’s delay screen proceeds from the event chronology to the affected workfront, then to a contemporaneous program showing the controlling path, and finally to concurrent events and the contract’s allocation rule. The first two steps can identify a plausible interruption; the last two are needed to attribute an extension of time or compensable disruption. In this case, the program path and concurrent activities remain to be established, so the study reports the disagreement and the missing link rather than a calculated delay. Evaluating an extension of time would require a dated activity-level program, progress updates, the affected workfront, and contemporaneous records of other delays, which would also allow concurrency to be tested before time or cost is attributed.
From these observations, the study derives a record sequence for claim administration. A change register should connect the dated requirement or instruction to a drawing revision, a quantity take-off version, and a valuation sheet. Each cost component should state its measurement basis and whether it has been requested, reviewed, agreed, or rejected. A separate time register should link the affected activity and workfront to schedule updates and record concurrent causes before an extension is quantified. The author notes that these controls follow the gaps observed in the three bundles, and that assessing their relative effectiveness across projects would require a multiproject design. As a procedural comparator, the study references the International Federation of Consulting Engineers’ 2017 Red Book, reprinted with amendments in 2022, which distinguishes claims from disputes and specifies clearer notices, while emphasizing that contractual entitlement in the Korean case depends on the actual agreement, instructions, notices, and applicable law.
The limitations are stated plainly. This is a single-project study of three retrospectively selected bundles from an accessible litigation exhibit set, and the full exhibit inventory and a systematic selection framework were unavailable, so eligibility and exclusion counts could not be reconstructed. Selection by availability may favor more complete or distinctive bundles, and the presented ratios describe these documents rather than a share of total project costs or an observed settlement outcome. Complete schedule updates, affected-path logic, and a full project event register would be required to measure delay, evaluate concurrency, and compare claim outcomes. The original exhibits contain identifying information, which also constrains public replication. What the study delivers instead is a transparent classification of where the evidence chain changes status: a coherent cost build-up that stops short of authorization, a quantity version mismatch that signals a measurement-basis difference, and a contested stoppage whose component shares reconcile even as the parties disagree about eligibility of the claimed cost and time. Replication with complete programs and additional projects, the author concludes, could then test how often these documentary gaps affect negotiated or adjudicated outcomes.
Subject of Research: A documentary analysis of construction claim records from a thermal power plant coal-handling package
Article Title: Tracing Construction Claims Through Cost, Quantity, and Work Stoppage Records in a Thermal Power Plant
Article References: Bae, B. (2026). Tracing Construction Claims Through Cost, Quantity, and Work Stoppage Records in a Thermal Power Plant. Architectural Research, 28(1), Article 25. https://doi.org/10.1007/s44515-026-00030-0
Image Credits: AI Generated
DOI: 10.1007/s44515-026-00030-0
Keywords: construction claims, thermal power plant, change orders, cost estimation, quantity take-off, delay analysis, contract administration, evidence chain, extension of time, coal-handling works, documentary records, project management
News Source: Denise Maddox. (October 27, 2026). Document Gaps Decide How Far a Construction Claim Can Be Proven. Scienmag.
