InsureMEP CriticalAsset

The Water Containment Report

A new standard for water risk.

What is installed? What feeds it? What can isolate it? How is a leak detected? Who can reach the valve, and what happens after the water is off? One building-specific assessment answers all of it — with the evidence attached.

Written and checked on CriticalAsset →

CARS · the CriticalAsset Risk Score

Score the controls. Show the evidence.

The Water Containment Report scores a building with CARS — the CriticalAsset Risk Score — across eleven weighted categories, each scored against the building’s own records. The overall CARS score and band are recomputed by a deterministic checker from the category scores; the arithmetic is never left to a model. A category with nothing on record scores nothing, and the report is marked provisional until the evidence is captured.

  1. Asset graph completenessIs every water asset on record?
  2. Fed-from / fed-to qualityDo we know what feeds what?
  3. Isolation controlCan the right valve stop the right problem?
  4. Detection coverageWhere would a leak be noticed — and where wouldn’t it?
  5. Power dependency controlWhat does the response need electricity for?
  6. Maintenance obligation executionAre inspections and service actually done?
  7. SOP & curtailment readinessIs there a procedure people can use?
  8. Recovery & restoration readinessPartners, spend authority, the route back to use.
  9. Physical access readinessKeys, doors, after-hours — the locked-door problem.
  10. Tabletop & training proofHas anyone rehearsed it?
  11. Time-series evidence qualityHow fresh, continuous and trustworthy is the data?

CARS bands

AA ≥85A ≥75BBB ≥62BB ≥50B ≥35CCC

What you get

A score that comes with its homework.

  • A CARS score and bandCategory by category, recomputed and checked before it’s shown — and flagged provisional while evidence is thin.
  • Evidence, partial, or nothing on recordEvery category shows what stands behind its score. Uncited judgement is labelled as judgement.
  • Single points of failureThe valve that, if it fails or can’t be reached, takes the most of the building with it.
  • Defects, not guessesImmediate, near-term or planned, with an estimated cost. A hypothesis is labelled a hypothesis and never moves the score.
  • Tiered corrective actions — and where they take the scoreWhat to verify, repair or rehearse first, with target dates and the projected CARS at 120 and 365 days.
  • An underwriting narrativeVerified controls, partially verified items and what is not evidenced — written for the broker’s office.
  • Every sentence citedAssets, documents, obligations and work orders. Invented references are rejected by the checker.
ExampleRevision 7Provisional
Water Containment Report Illustrative office tower 22 floors · illustrative example, not a customer
CARS43Bprovisional

Provisional. 25% of the score’s weight rests on categories with nothing on record. CARS gives no credit for what isn’t on record — the score moves when evidence is captured.

01 · At a glance

CARS · CriticalAsset Risk Score43/100 — Brecomputed by the guard
Requirements met4 of 5missing: plumbing plan
Evidence behind the 11 categories 4 evidenced · 4 partial · 3 with nothing on record
Open defects (evidenced)1 immediate 2 near termplus 1 hypothesis to verify
Corrective capex$18,650model estimate
Citations resolving12 of 12every one opens a record
Evidence sufficiency (MAX)5 sufficient 4 insufficient

02 · CARS categories

Asset graph completeness12.5%72partial · 96
Fed-from / fed-to quality12.5%64evidenced · 41
Isolation control12.5%58partial · 15
Detection coverage10%42evidenced · 18
Power dependency control7.5%76evidenced · 9
Maintenance obligation execution10%0nothing on record
SOP / curtailment readiness10%30partial · 2
Recovery & restoration readiness10%0nothing on record
Physical access readiness5%55partial · 6
Tabletop / training proof5%0nothing on record
Time-series evidence quality5%61evidenced · 2,840
Overall100%43 B

Categories with nothing on record score zero until evidence is captured. Narratives without a citation are labelled as judgement, not fact.

03 · Single point of failure

Riser 2 sectional valveisolates floors 12–20 · computed from the building’s own valve-to-line map, never taken from the model

04 · Open defects

Immediate
No named after-hours access to the B1 mechanical roomEvidenced: access register lists no keyholder
Near term
No flow meter or night-flow detection on the domestic mainEvidenced: the asset register holds no flow meter
Near term
Riser 2 sectional valve: no operability evidenceEvidenced: no exercise record in the valve log
Hypothesis
Abandoned-in-place piping on level 3Not scored — verify first. A defect moves the score only when the record evidences it.

05 · Corrective actions

43 today → ~61 in 120 days → 78 in 365 days (model projection)
ImmediateName and post an after-hours keyholder for the B1 mechanical room30 days$900
Near termInstall a flow meter on the domestic main; 30-day night-flow baseline and anomaly alerts120 days$9,800
Near termExercise riser 2 sectional valve; log operability evidence120 days$1,450
ProgrammaticQuarterly tabletop covering the undrilled water scenarios365 days$6,500

06 · Validation

47 → 43 BGuard verdict: corrected. The model claimed 47; the trained weights give 43. 43 is reported. Two sentences calling obligations “overdue” were withheld — the record holds no completion data, so they are unevidenced, not overdue.

CLEO composes the assessment · MAX judges whether cited evidence supports each control · the guard recomputes the score and checks every citation. Not a certification, and never an instruction to operate equipment. Every action lands for human review.

How it’s built

Facts first. Judgment second. Proof always.

1 · The building record

Plans, asset inventory, fed-from topology, documents, maintenance obligations and work history — assembled from what the building already has, and from a walk of the site where needed.

2 · Specialist assessment

Purpose-trained specialists read the record and produce the category scores, defects and corrective actions in a fixed, auditable structure.

3 · Deterministic check

A checker recomputes every score and band, verifies every citation and flags any sentence that reads like an operating instruction.

4 · A living document

When a valve, a document or a requirement changes, the report says the assessment is stale — and a re-run refreshes it.

Alignment review — not certification. The report states requirements; qualified people decide and carry out actions. No insurance outcome is promised. The report is delivered on CriticalAsset.

Find the delays before a leak does.

Start with the free Emergency Controls Report — your floor plans and emergency controls, mapped at no cost. The Water Containment Report builds on it.