How to document the cause of a building water loss

What belongs in a water-loss cause record?

Document a building water loss by linking the observed release, water path and affected materials to dated photographs, measurements and source tests in one evidence register. Keep observations, witness accounts and causal inferences separate, and record the conditions and limits of every test. Build a chronology that distinguishes discovery from onset, preserve the original evidence, and state which cause is supported and which alternatives remain unresolved.

Build the record around a chain: source, release or entry point, transport path, affected material. The source is the supply or environmental origin; the release point is where water escaped or crossed the enclosure. The failure mechanism explains why the component failed, while damage describes the resulting material condition. A leaking coupling can be established while its fracture mechanism remains unresolved.

IICRC's public description of ANSI/IICRC S500, fifth edition, 2021, assumes that the property owner arranges underlying source determination and correction and can contract with the restorer or a specialist for that work. Name that investigator and the assigned systems in the scope. This scope assumption is distinct from the restorer's initial source verification discussed in the 2026 position statement. IICRC S500 scope; 2026 position statement.

Use the mitigation and structural drying guide for extraction and drying work. Cause, contamination, drying completion and policy coverage answer separate questions; a good cause record gives each decision its own evidence. The register conventions below are Anvilfield reporting practices, not statutory burdens of proof.

What should you capture before cleanup changes the scene?

Record arrival, active discharge, isolation and the response as separate events. Capture accessible evidence while controlling damage; do not leave a release running to finish photographs. Mark inaccessible spaces and the reason access was restricted, so an uninspected cavity never becomes an assumed dry cavity.

Establish safe access before collecting evidence

The American Industrial Hygiene Association’s Flood Recovery: Hazards guidance identifies electrical, structural, sewage and chemical hazards, including lead and asbestos in damaged materials.

Keep people out of unstable or electrically hazardous areas; assign electrical isolation and verification to the electrician before entering wet work areas. Do not operate breakers while wet or standing on a wet surface. Before opening walls or ceilings, establish the controls for suspect materials and contamination, and record the access restriction and responsible trade in the register.

Photograph the as-found condition before each intervention: opening a ceiling, clearing a drain, moving a valve, applying temporary covering or removing a component. When emergency repair has already occurred, identify the repairer, request the original photographs and invoice, and locate removed parts. State which observations survive and which conditions the repair changed.

FEMA/NFIP FloodSmart documentation guidance calls for wide and close photographs before cleanup, appliance serial numbers, small material samples and repair receipts. Apply those evidence practices to the loss record while keeping the program’s flood-claim coverage terms separate.

EPA's Mold Remediation in Schools and Commercial Buildings, chapter 4, uses a 24 to 48 hour clean-water response window and explains that mold can develop before 48 hours or fail to develop afterward. That is a response guideline, not a timestamp for the release or an automatic contamination classification. EPA response guidance.

For retained parts, log component ID, installed orientation, removal time, remover, container or seal, storage location and every transfer. Preserve the failed surfaces: agree the examination before cleaning, cutting or testing the failure face. These are Anvilfield preservation practices; record any emergency handling that prevented them.

How do observations differ from causal conclusions?

An observation states what someone saw or measured; an inference connects evidence to a proposed explanation. Keep witness accounts attributed, either verbatim or clearly labeled as summaries. “Water appeared after rain” describes a reported sequence; it does not identify the opening through which water entered.

Start the register with property and loss IDs, investigator, scope, inspection dates and timezone, drawing revision, evidence-storage location and report revision. Every entry needs event or collection time, recording time, location or point ID, recorder or originating source, original-file reference and linked photograph or test IDs. The following entries demonstrate field wording; their IDs are separate from the municipal case later in this guide.

Use four conclusion labels: confirmed at the tested location and conditions; supported by converging evidence; unresolved; and not reproduced under recorded conditions. Apply each label to a proposition, such as release point, path, mechanism or onset. They are Anvilfield report language, not numerical probabilities or ASTM grades.

ASTM E2128-26 addresses exterior-wall leak evaluation and warns that sampled locations do not identify every leak. One confirmed release therefore supports a finding at that location; it does not establish a whole-building explanation. A confirmed present release can coexist with unresolved historical duration.

How do you build a defensible event timeline?

Separate last verified dry condition, earliest report, first direct observation, isolation, testing, removal, repair and retest. A photograph records the visible condition at capture; it does not establish when leakage began. A prior dry inspection constrains only the rooms, materials and layers actually checked.

Keep event time and recording time in different fields. Preserve device time, record timezone or mark it unknown, and log clock corrections without overwriting the original. A witness who remembers only the day supplies a date-only entry; adding an invented clock time makes the record less accurate.

Keep conflicting dates side by side with their source documents and a resolution action. Sensor alarms, building-management trends, utility intervals, occupancy records and repair history narrow the chronology when their clocks, sampling intervals and coverage are known. Staining, mold, corrosion and today’s flow rate cannot independently date onset.

NOAA NCEI’s Local Climatological Data Version 2 supplies station weather context. Retain dataset version, station ID and location, distance from the building, observation period, time convention, units and quality flags with the export. LCDv2 CSV data use metric units; retain the native values and document any unit conversion. Pair station rainfall, wind and freeze/thaw records with site observations to establish exposure at the actual opening.

Source: City of Covington, October 26, 2021 council packet. Gaps compare calendar dates, including the proclamation’s date rather than its clock time. Ask the record custodian whether the October 6 proposal revised the bid reported received October 5; retain both until the original transmittal resolves the sequence. Receipt, inspection and document-issue dates belong in separate fields.

Which test separates the competing sources?

Choose the test that changes the decision. Write the proposed source, controlled input, expected observable response and competing explanation before starting. Begin with records and direct observation, then nondestructive checks, controlled isolation or operation, and targeted access or wetting where needed.

Change one input at a time and account for simultaneous water uses. A positive meter check establishes flow through a boundary; observing the release establishes where water escaped. A plumbing defect and a roof defect can both contribute to the same loss.

Use the full applicable method and equipment instructions for execution. The supply row follows the EPA facility meter-check approach; the enclosure row draws on E2128-26 and E1105-26. Other rows are Anvilfield test-selection conventions, not standardized acceptance procedures.

Draw the actual supply boundary using the pressure, booster and PRV system guide. For drain evidence, retain access and distance references described in the CCTV inspection guide. Record cleaning before inspection because it changes the obstruction evidence.

For each controlled test, record the baseline condition, input start and stop, observers, first response and continuing drainage after the input stops. Distinguish entry at a defect from water already stored in insulation or a cavity. If a test wets several connected details at once, narrow the zone before naming one detail as the entry point.

Set stop conditions before intentional wetting: observed entry, threatened contents, uncontrolled migration or unsafe access. Record negative results with duration, resolution, inaccessible surfaces and operating conditions. Record repairs and depressurization as changed conditions; use an investigation-specific test procedure and pressure.

What do meters and thermal images actually prove?

EPA's October 2012 WaterSense commercial guidance uses a one-hour no-use meter check, its current household guidance uses two hours, and Seattle Public Utilities uses 30 minutes, so a cause record must identify the procedure and actual observation interval. Commercial procedure; household procedure; Seattle procedure.

Use EPA’s commercial procedure when facility demand can be stopped. Inventory toilets, irrigation, ice makers, softeners, automatic makeup and storage filling; log permitted loads individually. Retain meter ID, units, display increment, update interval, start/end readings and times, and a sketch identifying each valve’s boundary. Assign fire-suppression changes to the system maintainer.

FLIR’s moisture-detection guidance explains that infrared reveals surface-temperature patterns and calls for moisture-meter confirmation. Pair thermal and visible images; insulation gaps, reflections and air movement offer competing explanations for a temperature pattern. Record local air temperature and relative humidity alongside the surface conditions.

Establish a known-dry reference on matching material, finish, substrate and depth, following the Tramex ME5 guide, printed pp. 5 and 10. Mark every reading on the plan and increase mapping detail around junctions and changes. A different finish or concealed metal can explain a different response; do not automatically extend the wet boundary to every high reading.

Preserve the instrument’s native scale

References: ME5 user guide, undated publicly served edition and SurveyMaster BLD5375 sheet AAS-920-085G, July 2024. Retain model, serial number, mode, material selection, depth, instrument check and native reading. Resolve a failed instrument check before mapping and identify readings needing repetition. Never average incompatible relative scales into a moisture percentage; none of these readings independently dates wetting or identifies the source.

How do you distinguish plumbing leakage from rain intrusion?

Compare evidence combinations: metered flow without intended demand, response to verified isolation, exposure-linked wetting and an observed route from the proposed source. Rain timing and a ceiling stain alone cannot choose between plumbing and enclosure entry. Domestic-meter silence also leaves separately fed or stored-water circuits open.

Map joints, penetrations and connected layers, including lateral travel and capillary transport where supported. The stain locates water arriving in the visible room; the entry point can be elsewhere. Pressure change alone also needs explanation: temperature, trapped air and valve leakage can affect it. Connect the suspected branch to an observed release or corroborating route.

Separate supply operation from fixture discharge

A fixture cycle can wet a supply connector, overflow, drain joint or shower enclosure. Observe filling, standing water and discharge as separate stages with other uses logged. A downstream drain defect does not explain water appearing before discharge, while a quiet supply meter does not exclude leakage from water already held in a fixture or pipe.

For intermittent losses, return during the relevant appliance, heating, cooling or rainfall cycle and retain the operating log. Compare isolated and restored states only when restoration is safe and authorized; use photographs, repair records and circuit data when recreating the release would cause damage. State which original conditions could be recreated.

ASTM E2128-26 concerns exterior walls and excludes roof leakage, below-grade leakage and condensation from its scope. Use the roof diagnosis guide for roof inspection and the electronic leak detection guide to match detection to the roof assembly.

Keep diagnosis and performance testing distinct

ASTM E1105-26 §5.5 says not to conduct the water-penetration test when wind gusting causes pressure to fluctuate more than ±10% from the specified test pressure. Retain specified and measured pressure, not just a pass/fail label. ASTM E1105-26 public significance and use.

E1105 evaluates installed windows, doors, skylights and curtain walls under specified pressure and water application. An uncontrolled hose spray is not that test. For progressive roof wetting, prepare a separate zone map showing runoff, observer locations and the lag after stopping; select duration and application for the actual assembly and investigation.

Compare test exposure with loss exposure. A dry result under one pressure or wind direction leaves other conditions open; a positive result establishes entry under the test conditions. State separately how well those conditions represent the reported event.

What other sources and classifications must you check?

Keep condensate, drain surcharge, condensation, irrigation, groundwater and fixture overflow in the source list. More than one source can be active. Follow the condensate drain and trap guide through the pan, trap and discharge route, and the subsurface drainage guide for exterior and below-grade pathways.

The National Weather Service dew-point training page defines dew point as the saturation temperature. With selected inputs of 15°C local dew point and a 12°C pipe surface, Anvilfield estimates a surface margin of -3°C. Subtract dew point from surface temperature, compare the margin with instrument uncertainty, then confirm actual moisture and its route.

This temperature relationship permits condensation at the cold surface; observed droplets and a connected wetting path establish whether it contributed to the loss. Use the pipe insulation and condensation-control guide to examine insulation continuity and local exposure. Capture cold bridges during heating conditions, chilled surfaces in humid air and condensate during cooling cycles; record concealed surfaces that remain inaccessible.

Separate groundwater seepage from drain surcharge by recording sump levels and operation, exterior water levels, fixture responses and the first observed entry points. A camera-visible pipe crack identifies a defect; tie it to the loss through flow conditions and the escape route. Inspect irrigation timing and nearby discharge points when wetting occurs in dry weather.

The IICRC S500 Consensus Body's February 10, 2026 position statement says that top-down entry and wind-driven rain do not automatically make a loss Category 3. Determine contamination from inspection and information gathering under S500-2021, keeping that evidence separate from the cause finding. IICRC weather-related position statement.

How should photographs and retained evidence be indexed?

Index photographs by location and finding. Anvilfield’s practical sequence is a locating view, assembly view and detail where needed; include unaffected references and instrument mode. Locate each joint by room, orientation and connected pipe.

The assignments below specify views to capture. Each completed entry also needs capture time and timezone, recorder, original filename, original-file location, hash and transfer log. Match the IDs to the observation register and plan; preserve the same point identifiers through opening, repair and retest.

Caption the component, visible condition, direction and linked finding; index video by segment time and file ID. Photograph parts before removal and after labeling, and log alterations made to stop the loss.

NIST IR 8387, September 2022, §3.2 supports preserving source and transfer information, retaining original digital evidence and securing hashes separately. Apply those practices to photographs, videos, meter exports and reports; annotate copies and log revisions. A hash establishes file identity against the recorded hash, not scene accuracy or capture time.

Hand over a folder index connecting the report and revision log to plans, registers, timeline, originals, raw readings, test sheets, witness sources, part transfers and open actions. Preserve metadata before resizing images. Agree who retains the originals, who can access them and the retention period in the scope; a building-loss record has no single retention period applicable to every contract and jurisdiction.

Worked example: turn a public loss record into bounded findings

The Covington police-building proclamation initially names roof and fire suppression as alternatives; later city accounts attribute the leak to roof defects after contractor photographs. The attached proposal addresses vent flashings, ridge-cap seams and possible panel defects. Proposed repair locations and verified entry routes remain different propositions. Covington council packet, printed pp. 72, 77 to 80.

Worked cause statement: At the Covington police building, the city attributed recurring ceiling water to roof defects after reviewing contractor photographs reported received October 5, 2021 (C-01 to C-03). The city’s roof attribution is supported by its account; the entry-to-ceiling path, disposition of the fire-suppression alternative and repair retest remain open in this reconstruction. The attached proposal bears October 6 (C-04), requiring reconciliation with the receipt account.

Method: Anvilfield calculates the elapsed calendar days in the table by subtracting discovery from each named record date. Keep the receipt account and proposal date as distinct endpoints pending reconciliation. Establish actual wetting duration from onset, intermittent operation and cessation evidence.

Estimating gallons requires a rate history and an operating duration

EPA’s October 2012 WaterSense commercial guide publishes a broken-line scenario of 15 gallons/minute for one day, totaling 21,600 gallons. For that selected rate operating for 12 hours, Anvilfield estimates 10,800 gallons. EPA WaterSense at Work §2.3.

Method: Multiply the rate by minutes of operation: volume = flow × time = 15 gallons/minute × 720 minutes = 10,800 gallons. For changing rates, sum the supported operating intervals.

Establish the rate from a timed measurement, meter interval or applicable equipment record, and record pressure and operating state with it. Use separate low and high scenarios when onset or rate is bounded rather than measured; keep the selected duration beside each result.

When must the cause remain unresolved?

Leave a proposition unresolved when its source path, exposure or competing explanation lacks adequate evidence. Assign the next action and its owner; unresolved is an actionable finding. Use this cause-statement sequence: location, supported source, observed or tested path, supporting IDs, conditions, and remaining mechanism question.

Specify whether the statement describes the loss event, inspection state or retest. Give each alternative a disposition: supported, contradicted under stated conditions, or still open. A dry fair-weather visit after roof repair does not address the original wind-driven rain exposure.

For burst-versus-gradual questions, separate release, duration and mechanism. Use prior incidents, occupancy, alarms, local temperature records and part examination where relevant; a split pipe alone does not establish freezing. When a failed part was discarded, retain the repairer’s photographs and attributed account, and identify the examination that is no longer possible.

Close the record with source-control status, repair evidence, a relevant retest and retained alternatives. Reopen it when water recurs, a conflicting original emerges or a relevant operating condition changes; preserve the previous conclusion and revision reason. Hand over completed work and assigned open actions even when historical cause remains unresolved.

Unresolved-cause checklist: Mark yes, no or not applicable; every no receives an action, owner and due date. This is Anvilfield’s completeness check, not certification to a standard.

• Property, room, level, component and loss identifiers match throughout the file.
• Discovery, last verified dry condition and possible onset are separate entries.
• Witness statements identify source and timing precision.
• Conflicting dates and clock corrections retain originals and resolution actions.
• Every causal inference links to specific evidence IDs.
• Supply, drain, roof/wall, HVAC, fire/process and below-grade alternatives have dispositions.
• More than one source has been considered.
• No-use demand inventory, meter boundary and valve positions are documented.
• Moisture readings retain model, scale, material, depth and matching control.
• Thermal indications have independent material checks.
• Weather records retain station, period and units; site observations remain distinct.
• Negative tests identify untested conditions and inaccessible surfaces.
• Cleanup, opening, isolation and repairs appear on the timeline.
• Failed parts and original files have traceable storage and transfers.
• Repair verification includes test conditions and observed response.
• Final conclusions identify scope, remaining alternatives and next actions.
• Every unresolved item has an action, owner, due date and closure evidence.

How much time should the documentation scope allow?

Anvilfield estimates six staff hours for the documentation tasks below, using selected allowances of one hour for records, two for mapping, one for test attendance and two for handover. This planning allowance covers an accessible area, one coordinated visit and one targeted test; travel, specialist testing, laboratory work, repair and return visits need separate allowances.

Separate staff hours, unattended observation and elapsed days. Meter observation takes the source-specific interval described above; access, weather and equipment availability determine the calendar. Schedule returns during the relevant exposure or operating cycle.

Method: Anvilfield estimates sum the four selected staff-hour inputs dated November 5, 2026. Add task hours for each additional area, system, test and return visit, and track actual hours against those tasks. Identify which tests require a separate trade mobilization rather than treating test attendance as test execution.

Scope initial preservation, mapping, targeted tests, original-data handover, cause statement, repair verification and ownership of unresolved items as separate deliverables. Name the owner-appointed investigator and trades authorized to isolate or test each system. Separate emergency source control from examination of the underlying failure mechanism, and assign each deliverable an acceptance record.

Restoration estimators pursuing school and municipal response contracts find public restoration and cleaning bids and early project signals from budgets and board meetings through PorchRocket Bid Control.

Sources, review date and citation

Byline: Anvilfield. Reviewed November 5, 2026. All primary references below were accessed on that date. Public standard scopes support the applications described here; the full applicable method governs test execution.

1. IICRC, ANSI/IICRC S500 publisher page, fifth edition, 2021; restoration and source-investigation scope.
2. IICRC S500 Consensus Body, Category of Water Damage and Weather-related Events, February 10, 2026.
3. ASTM E2128-26, Standard Guide for Evaluating Water Leakage of Building Walls; public scope and significance.
4. ASTM E1105-26, field water-penetration testing of installed exterior assemblies; public scope and significance.
5. EPA, Mold Remediation in Schools and Commercial Buildings, chapter 4, current web presentation; response guidance.
6. EPA, WaterSense at Work §2.3, Leak Detection and Repair, October 2012 text hosted May 2023.
7. Tramex, Moisture Encounter ME5 User Guide, publicly served undated edition, printed pp. 4, 5, 8, 10, 13 and 20.
8. FLIR, How to detect a water leak with thermal imaging, current undated web guidance.
9. EPA WaterSense, Fix a Leak Week, current household meter-check guidance.
10. Seattle Public Utilities, Detect Leaks, current undated meter and isolation guidance.
11. National Weather Service, Temperature: Dry Bulb/Wet Bulb/Dew Point, undated training page.
12. FEMA/NFIP FloodSmart, Document Flood Damage, current undated documentation guidance.
13. Amphenol/Protimeter, SurveyMaster BLD5375 product sheet, AAS-920-085G, July 2024.
14. NOAA NCEI, Local Climatological Data Version 2, current product and access documentation.
15. City of Covington, October 26, 2021 City Council Regular Meeting Agenda Packet, printed pp. 72, 77 to 80.
16. NIST IR 8387, Digital Evidence Preservation: Considerations for Evidence Handlers, September 2022, especially §3.2.
17. American Industrial Hygiene Association, Flood Recovery: Hazards, current undated safety guidance.

Update log, 2026-11-05: Reopened all primary references; verified standard editions, manufacturer scales, meter intervals and municipal chronology. Added safe-access and staged-test guidance; clarified condensation and roof-test findings; repaired rendered links and replaced price allowances with a documentation scope worksheet.

How to cite this page: Anvilfield, How to document the cause of a building water loss, 2026, https://anvilfield.com/field-guides/plumbing/water-loss-cause-evidence-register/. Append the section fragment and review date when citing a table.

Related tools

FAQ

People also ask

• What is water mitigation?
• Where do most roof leaks come from?
• What is a sewer camera inspection?
• How thick should pipe insulation be?
• What is a pressure reducing valve?