A terracotta panel decision register turns early design assumptions into named questions, evidence tasks and approval gates. It prevents a promising material attribute from becoming an unverified building benefit in the brief, specification, budget or programme.
Use this workflow from concept to handover. Record what is known, what is provisional, who owns the answer, what evidence is due, which decision depends on it and what must be reopened when the product or wall design changes.
Put each statement at the level that can prove it
Start every register row by naming its evidence object. One document may inform several levels, but it should not be stretched beyond its stated specimen or assembly. If the proposed outcome cannot yet be assigned to a level and an owner, keep it open rather than turning it into specification text.
Panel
Exact body, profile, dimensions, tolerances, mass, finish, batch, edges and declared product results.
Components
Clips or carriers, rails, brackets, anchors, fasteners, gaskets and their compatible, traceable configuration.
Full wall
Cladding, cavity, insulation, membranes, backing wall, fire stopping and complete structural/enclosure behaviour.
Interfaces
Corners, openings, base, parapet, movement joints, penetrations and transitions to adjacent materials and systems.
Project outcome
Appearance, compliance, energy, comfort, programme, cost, maintenance, service planning and environmental goals.
For example, a product report can describe the tested panel specimen; it cannot by itself prove building energy use, whole-wall fire safety, installed cost or maintenance frequency. A project simulation or tender comparison is meaningful only when its wall build-up, interfaces, quantities, location, programme, study period and assumptions are visible.
Keep assumptions, requirements, evidence and decisions separate
| Record type | Meaning | Minimum control |
|---|---|---|
| Assumption | A provisional value or proposition used to keep design work moving | Source, confidence, owner, expiry date, dependencies and consequence if wrong |
| Requirement | A project target or mandatory condition | Metric, unit, boundary, acceptance criterion, applicable rule and approving party |
| Evidence | A calculation, report, declaration, sample, mock-up, inspection or other traceable record | Exact product/assembly, method and edition, author/lab, date, scope, result, limitations and current validity |
| Decision | An authorised choice made from requirements, evidence, risk, cost and programme | Decision maker, date, accepted configuration, conditions, residual risks and linked records |
A supplier statement can be logged as an input without becoming an approved decision. Likewise, a signed visual sample can approve a colour range but does not approve support capacity, cavity drainage, fire stopping or a future replacement match.
Create one traceable row for each material decision
The register can live in the project’s document-control platform, specification tracker or design-risk log. Use stable IDs and link rather than copy source records, so revisions remain visible. Avoid long narrative cells: a row should expose the next decision and the missing evidence quickly.
| Field | What to enter | Control question | Close-out condition |
|---|---|---|---|
| ID and topic | Stable number plus one precise question | Is this one decision, or have several claims been bundled together? | All linked records use the same ID |
| Evidence level | Panel, component, full wall, interface or project outcome | Can the proposed evidence prove a statement at this level? | Scope matches the claim and proposed configuration |
| Requirement | Metric/criterion, boundary, rule, project location/use and source | Is the target mandatory, client-selected or only desirable? | Requirement and approving authority are confirmed |
| Current assumption | Provisional answer, source, date, confidence and expiry | What work relies on it, and what happens if it is wrong? | Replaced by accepted evidence/decision or formally withdrawn |
| Owner and contributors | One accountable owner plus required reviewers | Who obtains, interprets and accepts the evidence? | Named people/organisations have completed their actions |
| Evidence due | Document/sample/mock-up/calculation needed, issuer and gate date | Is a current exact-product or exact-assembly record required? | Record is received, reviewed, linked and dispositioned |
| Decision and status | Open, at risk, accepted with conditions, rejected, superseded or closed | Are conditions and residual risks visible? | Authorised decision and date recorded |
| Change triggers | Product, factory, format, finish, support, build-up, interface, exposure, code, quantity or programme change | Which related rows and approvals must reopen? | Impact review complete and dependent records updated |
- the exact product and, where relevant, complete assembly are identifiable;
- the requirement and evidence use compatible scope, units, methods and acceptance boundaries;
- deviations between tested/calculated and proposed construction are assessed;
- interfaces and responsibility gaps are resolved rather than excluded silently;
- the decision maker, conditions, residual risks and next verification point are recorded; and
- dependent drawings, specifications, schedules, cost plans and programme information are updated.
Seed the register with questions, not promised answers
The examples below are starting rows. They do not certify a product, build-up or outcome. Project teams should split, combine or reassign them to match procurement route, jurisdiction, building use, height, exposure and design responsibilities.
| ID / level | Decision question | Evidence route | Gate / trigger |
|---|---|---|---|
| P-01 / Panel | Which current product, factory, profile, finish and size range can deliver the design intent? | Current data, drawings, declared intended use, physical range samples and manufacturing review | Before module freeze; reopen for item/factory/format/finish change |
| C-01 / Components | What retention/support configuration completes the load path to the verified substrate? | Coordinated system drawings, calculations, component data, tolerances and responsible-engineer review | Before detailed release; reopen for geometry, support, anchor or substrate change |
| W-01 / Full wall | What complete build-up is being assessed for structural, water/air, fire, thermal/moisture and acoustic requirements? | Project-specific calculations, test/assessment records, drawings and code/standards register | Before performance approval; reopen for any layer, cavity, barrier or joint change |
| I-01 / Interface | How are continuity, movement, drainage, support and fire strategy resolved at boundaries and penetrations? | Coordinated details, interface responsibility matrix, mock-up and inspection/test plan | Before construction issue; reopen for adjacent-system or opening-detail change |
| O-01 / Outcome | What appearance range and replacement strategy will be acceptable through procurement, installation and service? | Signed master/boundary samples, production control, elevation/batch map, cleaning trial and spare plan | Before production; reopen for batch, orientation, cleaner or replacement-source change |
| O-02 / Outcome | How does the compliant complete solution affect budget and programme against the defined alternative? | Same-scope quantities, quotes, preliminaries, mock-ups/tests, logistics, installation, access, maintenance and risk allowances | At cost/design gates; reopen for scope, market, quantity, sequence or programme change |
| O-03 / Outcome | What environmental conclusion is justified for the exact product and study boundary? | Current product/factory declarations and project comparison covering transport, construction, maintenance, service assumptions and realistic end of life | Before public or certification claim; reopen for source, plant, quantity, scenario or document-expiry change |
Move assumptions through five decision gates
Concept
Record intended outcome, candidate wall route, unknowns, comparison basis and high-consequence assumptions.
Schematic
Identify product families, outline build-ups, interfaces, responsibility, evidence routes, budget and programme ranges.
Developed design
Freeze coordinated geometry and requirements; review calculations, tests, details, samples and mock-up plan.
Procurement
Match submittals and quotes to the accepted configuration; expose substitutions, exclusions, qualifications and lead-time assumptions.
Delivery and handover
Close inspections, tests and nonconformances; link as-builts, batch records, maintenance, access, spares and residual risks.
A gate is not a date alone. Define which rows must be closed, which may remain conditionally open, who can accept the residual risk and what downstream work is prohibited until evidence arrives. If a critical assumption misses its gate, escalate the decision rather than quietly copying it into the next issue.
Compare equivalent solutions and declare uncertainty
A panel price, face-material mass or isolated product result is not a complete-option comparison. Compare functionally equivalent, regulation-compliant assemblies over the same facade zones, quantities, interfaces and study period. Include design and engineering, system components, substrates, mock-ups/tests, logistics, access, installation sequence, quality control, maintenance, replacement and realistic risk allowances. Record currency, quotation date, exclusions and sensitivity rather than publishing a generic “cost-effective” conclusion.
Apply the same boundary discipline to operational energy, acoustics, fire, water, programme, service planning and environmental comparisons. State the baseline, model or assessment method, location/exposure, inputs, limitations and responsible reviewer. Where evidence is not yet mature, show a range or scenario and keep the assumption open.
High-consequence uncertainty
Prioritise assumptions that affect life safety, compliance, primary load path, water/air continuity, procurement feasibility or irreversible interfaces, even if their probability seems low.
High-volatility uncertainty
Time-limit assumptions about current items, factory source, finish/size feasibility, quotations, lead times, standards, approvals and installer methods.
Linked uncertainty
Map dependencies. A format change can alter mass, support, jointing, fabrication, test applicability, quantities, appearance, logistics and programme simultaneously.
Residual uncertainty
Close a decision only when the remaining risk, owner, monitoring/verification point and contingency are explicit and authorised.
Reopen evidence when the configuration changes
At minimum, trigger review when the item or factory, clay body, finish, dimensions/profile, orientation, cut/edge condition, clips/carriers, rails/brackets, anchors/fasteners, substrate, cavity, insulation, membranes, fire barriers, joints, sealants/gaskets, interface detail, building use/height/location, code basis, quantity, installation method or maintenance plan changes. The proposer should issue a written comparison, identify affected register rows and show why existing calculations, reports, samples and approvals remain applicable—or provide replacements.
Maintain a decision history instead of overwriting superseded entries. Link revision-controlled drawings, specifications, calculations, test reports, declarations, samples, mock-up records, inspection records and nonconformance dispositions. At handover, convert active assumptions into verified as-built information, explicit owner actions or accepted residual risks; do not leave them disguised as generic benefits.
Start a project-specific terracotta decision register
Share the project location and use, facade zones, candidate panel or product family, outline wall build-up, interfaces, intended outcomes, applicable requirements, procurement route, comparison options and decision dates. Request current product/system evidence that can be linked to each open row.