As-Built Records and Design Reconciliation

As-built records describe the condition actually created by excavation, construction and installation. Design reconciliation compares that condition with the relevant design basis, approved changes and expected performance, then determines the engineering consequences of any difference. The objective is a dependable account of what exists, why it differs from the original proposal, and what that means for operation, future mining and subsequent modification.

Separate design intent, field evidence and engineering acceptance

The following working distinctions prevent a common ambiguity: an updated drawing can represent several different kinds of information.

The August 2020 Global Industry Standard on Tailings Management gives a mining-specific example of the breadth of an as-built record. Its Construction Records Report includes geometry, materials, test results, construction activities, quality records, changes, non-conformances and their resolution, photographs and instrumentation details. Requirement 6.3 calls for that report after a material change to the facility, its infrastructure or monitoring system, signed by the Engineer of Record and Responsible Tailings Facility Engineer. This tailings requirement illustrates why a final drawing alone cannot represent all the evidence about an engineered facility.^[1]

Status should remain visible within a record set. A surveyed pipe route can be accurate while its revised support arrangement remains under engineering assessment. An installation may also be complete while tests are outstanding. Collapsing these different states into a single “as-built approved” stamp hides the question that still needs an answer.^[2]

Make measured records comparable

Comparison begins with compatible reference systems and an identifiable observation date. A useful measurement record identifies the coordinate system, horizontal and vertical reference, units, control, method, coverage and quality assessment. The U.S. Geological Survey's Lidar Base Specification, 2025 revision A, requires survey and processing documentation covering control and checkpoints, reference frame, projection, units, vertical datum, geometric quality and acquisition anomalies. Those requirements apply to USGS lidar products; they provide a clear example of the supporting information needed to interpret surveyed surfaces.^[3]

For a mine, record any transformation between the site grid and the design model, including its provenance. Preserve areas with missing observations or low confidence. A dense point cloud does not by itself establish accuracy, and interpolation across an unobserved surface creates a modelled surface rather than a new observation. The uncertainty of the comparison should be considered alongside the tolerance being evaluated. These distinctions matter before a difference is classified as construction error, excavation overbreak or later movement.^[3]

Time is another reference. A later survey may capture settlement, blast damage or other changes. Record intervening events so differences are assessed against the represented state, without automatically attributing them to original construction or excavation.^[3]

Capture evidence before it becomes inaccessible

Records need to follow the work while the relevant features can be observed. Western Australia's August 2015 guide to TSF management and closure describes construction reports covering encountered foundations, material properties, changes from expectations, quality-control testing, rework and as-constructed survey drawings. It specifically includes buried services, drainage features and instrumentation. It also connects the construction report and monitoring data to the design of subsequent stages. These are technical record principles from that edition, rather than a statement of current submission obligations.^[4]

Applied to mine infrastructure, this supports planned capture before a drain is buried, reinforcement is covered, a foundation is loaded or an excavation becomes inaccessible. Link the observation to the feature and location it represents. A batch test demonstrates properties of the tested material; traceability is needed to associate it with the installed zone. Photographs need location, orientation, date and sufficient scale or context to answer the engineering question. Retain the raw evidence as well as its interpreted representation.^[4]

The capture plan should support future engineering decisions. For example, backfill records may identify placement stages and associated test samples. Select evidence that supports a design assumption or assessment, with traceability between the evidence and the installed feature.^[4]

Reconcile discrepancies through engineering judgement

A useful reconciliation follows the evidence from the observed difference to its consequence. First identify the correct approved revision and verify that the compared records describe the same location and state. Then distinguish a recording error, a measurement limitation, an authorised variation and an actual departure requiring assessment. Describe the discrepancy without altering the original observation to make it agree with the drawing.

ICMM's May 2021 Tailings Management: Good Practice Guide describes formal consideration of construction deviations, assessment of their effect on design intent and performance, and tracking of unresolved non-conformances. It also emphasises the cumulative effect of several minor deviations. Its construction-record guidance preserves the changes, their resolution and evidence of the constructed condition, including successive stages of facility construction. This provides a useful engineering pattern: a discrepancy is closed through an evidenced disposition, and the record remains an account of what was built.^[5]

Disposition may require correction, additional measurement, revised analysis or acceptance of a justified departure by the appropriate authority. The record should distinguish these outcomes. If an engineer accepts a revised arrangement, preserve its rationale and conditions; if work is corrected, retain evidence of the corrected condition. An unresolved departure should remain visible with its affected boundary and decision consequence.^[5]

Consider an illustrative case in which a drainage pipe is recorded above its intended elevation. Before changing the design model, establish whether the difference comes from datum handling, measurement or installation. If it is physical, the engineering assessment concerns the resulting drainage function and connected features, rather than the drawing discrepancy alone. If retained, the accepted arrangement becomes part of the baseline for later assessment. This example has no universal tolerance: the relevant limit depends on the design function and the reliability of the evidence.

Distinguish geometry, materials and operating performance

Geometric reconciliation compares locations, surfaces, clearances or volumes. Material and configuration reconciliation checks what was installed and the properties on which the design depends. Operational reconciliation compares measured response with the expected behaviour under stated loads and conditions. These questions can lead to different conclusions about the same installation: a component can occupy the intended position but have the wrong properties, or match its specified configuration while system performance remains unproven.

In excavated mine workings, Potvin and colleagues' 2020 benchmarking paper describes comparing cavity-monitoring surveys with planned stope shapes. Survey shadowing and the selection of the reference shape affect interpretation; the paper distinguishes comparison with a generic planned shape from comparison with the drill-and-blast design. It also explains how geological, geotechnical, blasting and survey information contribute to investigating overbreak and underbreak. A volume difference therefore identifies a result, while explaining its cause requires other evidence.^[6]

Production reconciliation introduces additional questions about tonnes, grade, stockpiles and processing. Those quantities may inform the interpretation of an excavation, but they are not interchangeable with a measured void. Keep the geometric comparison boundary and the material-accounting boundary explicit, particularly where development excavations or previously handled material could be counted twice. The stope study discusses the treatment of development volumes and the importance of retaining their positions when examining causes of deviation.^[6]

For mine fluid-power systems, NSW's MDG 41, Version 4 of April 2019, distinguishes installation verification, as-built records and commissioning. It recommends an inspection and test plan to check conformity and identify non-conformances, updated drawings and manuals, and quantifiable commissioning criteria. Installation and commissioning records should be retained in the plant safety file. These recommendations distinguish a documented configuration from evidence of functional performance.^[2]

An as-built record can faithfully document a non-conforming installation. Its existence therefore establishes neither fitness for purpose nor contractual or regulatory acceptance. Those conclusions need the applicable engineering assessment, testing and authorisation. When reviewing a pump circuit, for example, the recorded arrangement and the evidence of its specified pressure, flow and control functions answer related but separate questions. Preserve the link between the installed configuration and the conditions under which it was tested.^[2]

Use the reconciled baseline through the mine life

Handover should connect the record set with the information used to operate and maintain the asset. ICMM's 2021 guide calls for the operations, maintenance and surveillance manual to reflect constructed conditions and construction deviations, including relevant interfaces with the processing facility. A staged record should also remain understandable as part of the complete facility.^[5]

The baseline then becomes a reference for interpreting later monitoring. GISTM requires monitoring suited to checking design assumptions and identifying departures from expected performance, with review and design updates where necessary. Its change-management requirements also address the cumulative implications of alterations. An as-built baseline is consequently a starting condition for continued engineering assessment, rather than the end of the design evidence.^[1]

Maintain history distinguishing original condition, accepted alteration and subsequent deterioration. State evidence limitations and investigate them according to the decision being made. Reconciliation delivers value when findings change an assumption, improve a design, resolve a discrepancy or clarify the conditions under which an asset can be relied upon.^[1]