Live facility, new equipment, no acceptable downtime

    Data center retrofit commissioning software for upgrades where the existing IT load can never be the thing that fails.

    Capacity upgrades and retrofits add UPS modules, switchgear or cooling capacity into a facility that is already carrying live IT load. Deskely separates the new-equipment commissioning path from the live base build it's being added to, with method-of-procedure evidence and hold points that keep the existing critical load protected while the new capacity is proven.

    Scope
    New capacity in a live facility
    Constraint
    Zero impact to existing load
    Method
    MOP-driven hold points
    Gate
    Redundancy re-verified
    HALL 1 — CAPACITY UPGRADE ZONELIVE-SITE WORK PACK · REV DRETAINED LIVE LOADISOLATION ZONE 2 — PLANNED OUTAGEload transfer pointRETAINEDH1-Z02-ISORESTART GATED ON ISOLATION SIGN-OFF AND LOAD TRANSFER TESTTAG REGISTERUPS-B3New UPS module, tie-inRecords signedSTS-B1Static transfer switch, b…In testPDU-1.09Existing PDU, load unaffe…PreservationCRAH-04Replacement air handlerPunch open48 MOPsunder one system hierarchy

    The problem

    The new equipment is the easy part; proving it without touching live load is the hard part.

    Adding a UPS module, a switchgear section or a chiller to a facility already running IT load means every test step that could affect the live side needs its own method of procedure (MOP), its own hold point and its own rollback plan.

    When that evidence lives in a MOP document reviewed once before the work and filed afterwards, there's no record of which specific steps were actually executed, by whom, and what the live-side readings looked like at each hold point.

    Deskely turns the MOP into a structured, signed record — hold point by hold point — so a facilities manager can see exactly what happened during a maintenance window on a system carrying production load.

    MOP FILED, NOT RECORDEDUPS module addition — live busExisting IT load stays live throughout — every step is a risk to itEXISTING LIVE BUS — CARRYING IT LOADN+1 switchgear lineup, energised, in serviceMOP HOLD POINTS — PAPER PROCEDUREPre-check readings before switchingGo/no-go criteria confirmedLive-side reading during cutoverNot capturedSecond signature on breaker closureNot captured2 of 4hold points with no readings on file0acceptable gaps on a live IT loadEach hold point becomes a signed record with the readings that justified it.

    Setting up the register

    New equipment is modelled as its own subsystem, tagged against the live systems it will join.

    A new UPS module, switchgear lineup or chiller is registered as new scope with its own Level 1–4 commissioning path, while a reference to the existing live system it's being paralleled or tied into records the dependency explicitly.

    Single-line diagrams marked up for the change, along with the MOP itself, are parsed and linked so every switching step in the register cites the drawing and the procedure step it corresponds to.

    That link is what lets a reviewer confirm the new breaker's commissioning tests were completed before the MOP called for it to be closed into a live bus.

    1 · SOURCE DOCUMENTSAs-built single line rev…212 branchesMethod of procedure regi…48 MOPsNew equipment schedule18 assetsLoad transfer test plan9 transfersPARSE + MATCHtag · description · drawing ref2 · ONE TAG REGISTERUPS-B3New UPS module, tie-inHALL 1STS-B1Static transfer switch, b…HALL 1PDU-1.09Existing PDU, load unaffe…HALL 1CRAH-04Replacement air handlerMECHANICALTWO DOCUMENTS DISAGREEAs-built single l…Static transfer switch, boa…Method of procedu…Abbreviated descriptionEngineer picks the sourcedecision loggedWeeks of transcription become a draft an engineer reviews.

    Execution against the MOP

    Every hold point is a signed record, not a line ticked off on a printed procedure.

    Pre-checks, go/no-go criteria, live-side readings before and after each switching step, and rollback triggers are captured as structured fields rather than free text, so a hold point can't be passed without the readings that justify it.

    Where two people are required to authorise a step affecting live load, the record can require both signatures before the next step unlocks.

    Redundancy is re-verified once the new equipment is in service — an N+1 lineup that briefly runs N during the cutover has to be proven back to full redundancy before the maintenance window closes.

    MOP HOLD-POINT LADDERUPS module addition — live busPre-check readings, live side baselineSingle sign-offIsolate and prove dead, new UPS baySingle sign-offParallel new module, live-side readingDual sign-offClose breaker into live busDual sign-offRe-verify N+1 redundancyCxA sign-off2of 5 hold points cleared this window0steps unlocked without required readingsThe next hold point cannot open until the current one is signed, live-side readings included.

    Close-out

    The upgrade is signed off with an audit trail the next outage review can use.

    The completed MOP record, new equipment's own commissioning results and the redundancy re-verification form one dossier for the upgrade, distinct from but linked to the facility's base build history.

    Punch raised during the work is categorised against whether it affects the live side or only the new equipment, so nothing ambiguous is left open against systems carrying production load.

    If an incident review is ever needed, the exact sequence of switching steps, readings and signatures for that specific maintenance window is retrievable directly, not reconstructed from memory.

    AUDIT QUERYUPS-B3Hall 1 · UPS tie-in · new moduleFull history retrieved in one queryINSTInstallation checklistUpgrade contractor12 AugPFCPre-functional checklistCxA03 SepPARAParallel run recordVendor engineer18 SepMOPMOP sign-offFacility operationspendingCERTIFICATELTLive load transferpendingFacility operations managerPrerequisites demonstratedThe transfer certificate cites the exact MOP number and every prove-out record it depended on.

    Brownfield IST proves the whole facility again

    A capacity upgrade doesn't just add a UPS module, it changes the failure behaviour of every rack the shared bus already serves.

    Adding capacity into a live power train means the N+1 or 2N topology the facility was originally certified against no longer exists as tested — a new UPS module paralleled onto an existing bus changes load sharing, fault current and transfer timing for equipment that was never touched during the upgrade.

    Where the scope justifies it, a brownfield integrated systems test re-runs a scoped set of failure scenarios against the changed topology — a simulated utility loss or a forced transfer — rather than assuming the original IST result still applies to a bus that now behaves differently under fault.

    Deskely keeps that re-test scenario matrix linked to both the new equipment's commissioning records and the facility's original IST results, so a reviewer can see exactly what changed, what was re-proven, and what the facility's redundancy claim rests on after the upgrade rather than before it.

    EVERY HOLD POINT, SIGNED IN SEQUENCEMOP — UPS module additionThe next step doesn't unlock until the current one is signed1Pre-check: live bus readings baselinedLive-side readings on file2 signatures2Go/no-go criteria confirmedLive-side readings on file2 signatures3Breaker closed into live busLive-side readings on file2 signatures4Redundancy re-verified to N+1Live-side readings pendingPending3 of 4hold points signed with live readings0steps unlocked without the prior sign-offFacility stays back to N+1 before the maintenance window closes.The MOP is a structured, signed record — not a printed procedure filed after the fact.

    Evidence

    The MOP-verified dossier

    New capacity landing inside a live facility is proven twice: once that the addition works, and once that the existing load never saw it happen. Every hold point in the method of procedure needs its own signed record.

    RECORDS PRODUCED6 typesMethod of procedure (MOP) hold…Continuously through the worksPre-work redundancy baseline…Before any live switching or tie-inBattery impedance baseline recordImmediately before the upgrade beginsLoad bank step-load test report…During commissioning of the new…Post-work redundancy…Immediately after energisation of the…SLA acceptance and witness packFinal gate closing out the worksDOSSIERIndexed persubsystem and tagExport readyEach record lands in the same structure, so the dossier assembles itself.

    Method of procedure (MOP) hold point sign-off log

    Each step of the live-facility work was executed and verified in sequence before proceeding to the next

    Continuously through the works

    Pre-work redundancy baseline report

    The existing N+1 or 2N configuration and load levels immediately before the works are documented

    Before any live switching or tie-in

    Battery impedance baseline record

    Existing UPS battery strings are within acceptable impedance range before new load or capacity is added

    Immediately before the upgrade begins

    Load bank step-load test report for the new capacity

    The added plant accepts its rated load in steps without disturbing the existing critical load

    During commissioning of the new equipment, isolated from live load

    Post-work redundancy re-verification record

    Full redundancy is restored and confirmed at the shared plant level after the new capacity is integrated

    Immediately after energisation of the new plant

    SLA acceptance and witness pack

    The upgraded facility meets its uptime commitments with zero recorded impact to existing tenant load

    Final gate closing out the works

    How it runs

    From a MOP on paper to a signed, hold-point-verified upgrade.

    Retrofit work on a live facility rewards the same rigour as new build, applied to a much smaller and much higher-consequence scope.

    1. 01

      Register the new equipment

      New UPS, switchgear or cooling capacity tagged as its own subsystem with a reference to the live system it joins.

    2. 02

      Digitise the MOP

      Pre-checks, go/no-go criteria and hold points structured as a signed record tied to the switching sequence.

    3. 03

      Commission the new scope

      Level 1–4 testing on the new equipment completed and signed before the MOP calls for it to close into a live bus.

    4. 04

      Execute hold point by hold point

      Live-side readings captured before and after each step, with dual sign-off where the procedure requires it.

    5. 05

      Re-verify redundancy

      Confirm the facility is back to its designed N+1 or 2N state before the maintenance window closes.

    6. 06

      Close out and link

      Upgrade dossier issued and linked to the facility's ongoing base build record, punch categorised by live-side impact.

    FAQ

    Questions about Retrofit and capacity upgrade scopes.

    Ready to digitize
    commissioning and
    handover?

    We use cookies to improve your experience.
    You can opt out of certain cookies.
    Find out more in our privacy policy.