Leak detection cable zone maps and alarm tests before white-space turnover

Direct answer

Before white-space turnover, a leak detection cable zone map and alarm point test record should identify the room, data hall, white-space boundary, raised-floor or slab condition, equipment rows, cooling units, CDUs, chilled-water or glycol lines, condensate sources, humidification piping, roof or wall leak risks, drains, sprinkler or fire-protection exposure, leak detection controller, cable type, point sensors, leader and jumper cables, branch connectors, end terminators, installed cable length, zone names, reference distances, alarm points, BMS or DCIM integration, relay contacts, notification path, simulated leak test locations, response time, reset method, photos, exceptions, corrections, retests, witnesses, and the exact turnover release boundary.

The goal is to prove that the installed leak detection system matches the room, that the map can lead an operator to the leak, and that each required alarm point reports where the owner expects it to report. A coil of sensing cable under the floor does not prove coverage. A normal controller screen does not prove BMS mapping. A cable route sketched on a construction drawing does not prove the as-installed path after cabinets, condensate drains, pipe changes, and floor tile work are complete.

Do not invent a universal cable spacing, alarm delay, distance tolerance, water volume, sensor type, relay action, notification rule, reset sequence, alarm priority, or white-space acceptance rule. The manufacturer instructions, approved design, owner project requirements, commissioning script, controls matrix, BMS/DCIM standards, operations procedure, AHJ direction, and site safety plan control those decisions.

Why this record matters

Data halls contain equipment that can be affected by a small amount of liquid in the wrong place. nVent data-center leak detection material describes water, cooling equipment, air-conditioning failures, weather, chilled-water or refrigerant piping, condensate, roof or window leaks, chemicals, and generator fuel as leak sources that can matter to data center operations. ASHRAE's data center resources likewise treat cooling, humidity, liquid cooling, and technology-space operations as data center concerns.

The handoff problem is not only whether the controller is powered. The handoff problem is whether an operator can understand the alarm, find the leak, know which tile or equipment row to inspect, and know whether the signal reached BMS, DCIM, security, NOC, or another monitoring path. nVent's TraceTek application guide describes system maps mounted near alarm modules and based on as-built drawings. RLE and Vertiv manuals describe reference maps, zones, distance readings, notification methods, alarm history, dry contacts, and BMS or NMS integration.

The turnover packet preserves evidence at the moment the white space is being accepted. After racks, cable trays, power whips, containment, access flooring, and operations activity arrive, it becomes harder to prove where the cable was originally routed and whether a missed alarm was an installation problem, mapping problem, operations problem, or later damage.

Start with the accepted basis

Begin the record by naming the accepted basis: leak detection submittal, manufacturer installation manual, controls point list, BMS or DCIM integration matrix, commissioning test script, white-space turnover checklist, owner operations standard, as-built drawings, raised-floor tile grid, equipment layout, cooling piping layout, fire-protection plan, and safety plan.

TIA's ANSI/TIA-942 page describes a data center infrastructure standard with physical infrastructure scope, including architectural, electrical, mechanical, fire safety, telecommunications, security, and other requirements. NFPA 75 is an official standard page for protection of information technology equipment. Those sources are context for why data center physical systems and protection measures belong in a structured turnover record; the actual cable routing and alarm test acceptance still come from the project and product documents.

If the field team finds a conflict between the design map and the installed space, write the conflict. Examples include a CDU added after the drawing, a condensate route moved, cable hidden by floor pedestals, a rack row shifted, a sensor installed in a drip pan not shown on the map, or a BMS point name that no longer matches the controls matrix.

Map leak sources first

A useful zone map starts with likely leak sources, not with available cable length. Identify CRAC or CRAH units, CDUs, rear-door heat exchangers, chilled-water or glycol headers, valve packages, flexible hoses, humidifiers, condensate pumps, condensate drains, floor drains, roof leak paths, exterior walls, sprinkler piping, preaction rooms, mechanical chases, water service columns, janitor or restroom adjacency, generator fuel areas if in scope, and any owner-identified critical equipment.

nVent's computer-room application guidance lists leak sources such as air-conditioning units, chilled-water piping, plumbing and condensate lines, clogged drains, sprinkler systems, and building leaks. The nVent data-center whitepaper adds that point sensors can work when the leak point is defined, while cable sensors can help when the leak path is less predictable and location matters.

Record the reason for each protected area. A zone named Zone 3 is weak by itself. A zone named Hall A Row 14 chilled-water valve package to CDU-2 condensate drain tells operations why the cable exists and where to send the first responder.

Record the as-installed cable route

The cable route entry should include the controller, leader cable, sensing cable segment IDs, non-sensing jumpers, branch connectors, end terminators, point sensors, panel or module address, start point, end point, actual route, floor tile coordinates, equipment row references, column lines, wall landmarks, underside-of-floor obstructions, overhead exposure where used, and the installed cable length or calculated length used by the controller.

RLE installation guidance tells installers to follow the configured map, secure the cable with clips, keep the floor clean, avoid damaged or dirty cable, keep the cable in contact with the floor, use caution tags for distance references, and test the cable with the controller before and during installation. Vertiv's LP6000 manual similarly describes laying the cable according to the layout diagram, keeping it clean, maintaining floor contact, correcting the diagram to represent the as-installed route, and using that diagram for mapping and testing.

Do not let the route record stop at installed per plan. The field record should say whether the route changed, where it changed, who accepted the change, whether the map was revised, and whether the controller or distance calibration was updated after the change.

Build a zone and reference map

The map should allow an operator to move from alarm text to physical location. Include the data hall name, room boundaries, tile grid, rows, cabinets or reserved cabinet ranges, cooling units, CDUs, leak source labels, cable route, point sensors, distance marks, branch breaks, non-sensing jumpers, end terminations, module locations, controller location, BMS or DCIM point names, and the nearest safe access route.

nVent guidance describes a system map near the alarm module showing the alarm module, cable connections, landmarks, protected equipment, leak sources, and actual cable distances at connections or mapped points. RLE's controller material describes loading a reference map, marking coordinates, using zones, and showing leak location on a reference map. Vertiv's LP6000 material describes mapping leak-detection points after installation and measuring critical locations along the cable route.

A zone should not be larger than operations can use during response. If one physical cable covers several areas, divide the record into practical response zones, then make clear whether the controller, BMS, or DCIM also uses those same zones. Do not imply digital zone separation where the hardware only reports one summary alarm.

Test cable and point sensors

The test record should state the test method, test location, expected alarm, observed alarm, distance or zone shown, delay or response time where required by the script, reset method, returned-to-normal status, and whether the test reached local controller, relay, BMS, DCIM, email, SNMP, BACnet, Modbus, or other notification paths.

RLE guidance describes testing sensing cable with a small puddle or a clean damp cloth and warns not to saturate the cable because it can take longer to dry and reset. Vertiv's LP6000 manual describes touching cable with a clean wet cloth to verify alarm operation, drying the cable, then calibrating, mapping, and testing. Use the manufacturer method for the installed product and the approved commissioning script for pass or fail criteria.

Point sensors need their own record. Vertiv's LT410 point sensor guidance describes single-point leak detection for locations such as chilled water, glycol, humidification supply piping, condensate pumps and drains, overhead piping troughs, raised computer floors, and remote shelters, with Form-C dry contact outputs. A point sensor in a drip pan should be identified by pan, unit, alarm contact, and monitoring point, not merged into a cable-zone note.

Verify alarm points and integrations

List every alarm path the owner expects to use. Typical fields include controller local alarm, audible alarm, summary relay, leak relay, fault relay, cable break or contamination point, point-sensor dry contact, BMS point, DCIM point, NOC alert, security desk alert, email, SNMP trap, BACnet object, Modbus register, work-order trigger, and local annunciator.

RLE's LD2100 guide describes notifications through Modbus, BACnet, SNMP, SMTP email, or relay output, plus a summary Form-C relay with normally open, common, and normally closed contacts. Vertiv's LP6000 manual describes Modbus, BACnet, SNMP, SMTP, relay activation, Form-C dry contacts, leak relays, cable-break relays, alarm lists, and alarm history. Those sources support a record that separates local detection from remote notification.

For each point, record point name, address or object where available, normal state, alarm state, test method, expected message, actual message, receiving system, time received, acknowledged by, reset by, and discrepancy. A tested local leak alarm is not enough if the BMS point says general trouble, wrong data hall, wrong tile coordinate, or remains normal.

Show reset and fault evidence

A turnover test should not stop at alarm present. The record should show alarm acknowledgement, cleanup or drying method, reset action, returned-to-normal status, alarm history entry, fault status, cable break or end-terminator condition where tested, disabled notification restoration, and any latching or non-latching setting required by the script.

RLE's controller guidance explains that supervised systems can report leak, cable break, and cable contamination, and that latched alarms may require manual reset after the condition is gone. Vertiv's LP6000 material describes supervised leak and fault monitoring, alarm history, acknowledgement, reset, and drying the cable after a test. That makes reset evidence part of turnover, not an afterthought.

If the alarm did not clear, do not mark the area released. Record whether the cable remained wet, was contaminated, was damaged, lacked an end terminator, had a wrong setting, or required manufacturer support. A stuck alarm during turnover is an exception even if the original simulated leak was successful.

Use an auditable table

Use the commissioning form, owner turnover checklist, controls point-to-point sheet, or manufacturer startup record first. Add a field table where those documents do not connect the cable map, zone labels, alarm tests, photos, exceptions, and release decision.

Build a photo packet

The photo packet should show the controller, module labels, system map, installed cable route, cable caution tags or distance marks, branch connectors, end terminators where visible, point sensors, drip pans, cooling units, CDUs, valve packages, condensate pumps, floor tile coordinates, overhead exposure where used, BMS or DCIM screens, alarm history, and corrected deficiencies.

Use both wide and close photos. Wide photos locate the row, tile grid, wall, or equipment. Close photos prove the sensor, cable clip, tag, connection, point label, or screen value. If the route is under a raised floor, pair open-tile photos with a map reference so the reviewer does not have to guess where the tile was removed.

Keep failed photos. A missing end terminator, contaminated cable, cable lifted off the floor, wrong point name, active fault, or disabled notification should remain in the record with the correction and retest evidence. The final photo alone does not explain what was corrected.

Handle exceptions and partial release

White-space turnover can be partial. The release may cover Hall A Rows 1 through 8 while Row 9 remains held for map revision, or it may release local alarms while BMS integration remains held. The record should make that boundary obvious.

Common exceptions include cable route not matching the map, zone label not matching the floor grid, controller reporting the wrong distance, disabled relay output, BMS point normal during test, wrong alarm priority, no alarm history entry, alarm that will not reset, cable contaminated by construction debris, point sensor wired to the wrong input, or test location blocked by cabinets.

For each exception, record condition, location, expected result, actual result, correction, responsible party, retest method, retest result, and release decision. If the owner accepts a temporary watch, local-only alarm, or manual response procedure, record who accepted it and when it expires.

Before turnover checklist

Run this check before representing the leak detection cable map and alarm points as ready for white-space turnover.

• Confirm the release scope: data hall, room, rows, tile grid, cabinet range, controller, system ID, turnover milestone, and excluded areas.
• Confirm the accepted basis: submittal, manufacturer manual, controls matrix, BMS or DCIM standard, commissioning script, as-built drawing, owner requirement, and safety plan.
• Map leak sources: CRAC or CRAH units, CDUs, chilled-water or glycol lines, condensate, humidification, roof or wall exposure, drains, sprinkler piping, critical equipment, and adjacent water sources.
• Record the as-installed cable route: start, end, segment IDs, leader and jumper cables, branch connectors, non-sensing sections, end terminators, point sensors, route changes, and installed length.
• Verify the zone and reference map: zone names, tile coordinates, row labels, equipment landmarks, reference distances, module locations, map revision, and owner response path.
• Test sensing cable at representative mapped points using the manufacturer-approved method and the commissioning script; record expected zone or distance, observed display, response time, and reset result.
• Test point sensors separately with sensor ID, protected equipment, pan or location, dry contact, alarm input, BMS or DCIM point, observed message, and reset result.
• Verify alarm integrations: relay, Modbus, BACnet, SNMP, SMTP, BMS, DCIM, NOC, security, annunciator, and alarm history where required by the project.
• Record fault behavior and restoration: cable break or fault test where required, contamination status, end terminator, acknowledgement, latching or reset setting, and returned-to-normal evidence.
• Attach photos of maps, routes, tags, sensors, leak sources, controller screens, BMS or DCIM screens, alarm history, failed conditions, and corrections.
• Write every exception, hold, bypass, local-only alarm, monitoring limitation, map correction, retest, and partial release boundary.
• State the decision: released for white-space turnover, released except listed zones, released for local alarm only, held for correction, or retest required.

Weak and strong records

Weak note: Leak detection tested. Ready for turnover.

That note does not identify the data hall, cable route, zone map, leak sources, test locations, alarm points, BMS or DCIM signal, reset status, photos, exceptions, or release boundary.

Stronger note: Data Hall 2 leak detection system LDS-DH2 reviewed for white-space turnover on 2026-06-09 under commissioning script CX-LDS-04 revision 3, owner controls matrix CM-DC-22, and manufacturer instructions for the installed cable and controller. The release scope is Hall 2 Rows A through H, floor tile grid A1 through H38, with CDU-2B and the northeast roof drain chase excluded pending map revision.

The record includes the as-installed cable route from controller LDS-2 through leader cable LC-2, sensing cable segments SC-2A through SC-2F, two non-sensing jumpers at the main aisle crossing, branch connector BC-2, and end terminator ET-2. The zone map marks cooling unit CU-2-1 through CU-2-6, CDU-2A, condensate pump CP-2-3, chilled-water valve package VP-2-4, tile grid coordinates, row labels, and reference distances observed at each mapped test point.

Simulated leak tests were performed at six mapped locations using the approved damp-cloth method. The local controller displayed the expected zone and distance at each point, the BMS received leak alarm points LDS-DH2-Z1 through LDS-DH2-Z6, the DCIM alert showed the correct data hall and row, and alarm history entries were captured. Point sensor PS-CDU-2A alarmed locally and at BMS input AI-LDS-214. Point sensor PS-CP-2-3 reported the wrong message text as Hall 1 and remains held under issue LDS-019. All tested cable alarms were acknowledged, dried, reset, and returned to normal. Hall 2 Rows A through H are released for white-space turnover except CDU-2B, northeast roof drain chase, and PS-CP-2-3 remote message text.

The stronger note works because it names the basis, route, map, test points, receiving systems, reset evidence, exceptions, and turnover boundary. It does not turn one successful local alarm into complete system acceptance.

Common mistakes

The first mistake is treating the design drawing as the zone map. The turnover record needs the as-installed route and the map revision the owner will use.

The second mistake is testing only next to the controller. A useful test proves mapped leak points in the data hall, including far points, branch areas, point sensors, and locations near critical leak sources.

The third mistake is recording alarm present but not the message. The receiving system may show the wrong room, wrong row, wrong priority, wrong object, or no location.

The fourth mistake is ignoring reset. A system that alarms during the test but will not return to normal is not ready for turnover.

The fifth mistake is merging cable sensors and point sensors into one pass note. Point sensors have their own location, protected equipment, contact wiring, and alarm point.

The sixth mistake is releasing the whole white space when only local alarms were tested. Local detection, relay output, BMS, DCIM, notifications, and response maps are different evidence paths.

Questions that come up

How many leak test points are enough? Use the commissioning script, manufacturer guidance, owner standard, and risk map. The record should include enough points to prove each release zone, critical leak source, point sensor, remote end, branch, and integration path required by the project.

Can the system be turned over if BMS is not ready? Only if the owner and commissioning authority accept that limitation. The record should state local-only operation, temporary response procedure, expiration, and what remains held.

Should alarm points be normally open or normally closed? Use the approved wiring diagram, controls matrix, manufacturer instructions, and owner standard. The record should state the installed normal and alarm states, not choose them in the field.

Does a dry controller screen prove the cable is clean? No. It only shows current normal status. The route record should still document cable condition, floor cleanliness, contamination concerns, and photos where relevant.

Who signs the record? Follow the contract. Signers may include the electrical contractor, controls integrator, leak detection vendor, commissioning agent, data center construction manager, owner operations lead, BMS or DCIM representative, safety lead, and AHJ where applicable.

Compliance and safety limits

This field note is not a leak detection design, data center standard interpretation, NFPA interpretation, fire-protection approval, controls sequence, BMS/DCIM programming standard, commissioning script, alarm priority standard, manufacturer installation instruction, water-test procedure, energized-work permit, access-floor safety plan, white-space acceptance, or owner operations procedure. The approved project documents, manufacturer instructions, engineer, commissioning authority, owner, AHJ, controls team, operations team, qualified installers, and site safety plan control the work.

Do not use this checklist to bypass electrical safety, floor-tile handling, work-at-height controls, overhead piping controls, chemical or glycol handling, water release controls, lockout/tagout, alarm bypass approval, fire-protection restrictions, hot-work limits, access control, clean-room or white-space cleanliness rules, PPE, or qualified-person requirements. The packet preserves leak detection map and alarm-test evidence before turnover. It does not authorize unsafe testing or unapproved operation.

Sources checked

• TIA, ANSI/TIA-942 StandardUsed for official data center infrastructure standard context and physical infrastructure scope.
• ASHRAE, Data Center ResourcesUsed for official data center cooling, humidity, liquid-cooling, energy, and technology-space operations context.
• NFPA, NFPA 75 Standard DevelopmentUsed for official information technology equipment protection standard context.
• nVent RAYCHEM TraceTek, Sensing Application GuideUsed for computer-room leak sources, sensing cable layout, BMS connection, alarm relays, system maps, landmarks, cable distances, and mapped leak location context.
• nVent RAYCHEM TraceTek, Data Centers with Leak DetectionUsed for data-center leak source categories, BMS integration, point versus cable sensor context, leak location importance, response-plan context, and raised-floor/server-rack examples.
• RLE Technologies, SeaHawk Sensing Cable Installation GuidelinesUsed for sensing cable pretest, damp-cloth or small-water test method, cable cleanliness, floor contact, clip placement, caution tags, distance references, and map use.
• RLE Technologies, LD2100 User GuideUsed for controller distance reading, reference maps, zones, alarm history, relay outputs, Modbus, BACnet, SNMP, SMTP, reset, cable fault, and notification context.
• Vertiv, Liqui-Tect 410 Point Leak Detection Sensor Installation GuideUsed for point leak detection sensor applications, chilled water, glycol, condensate, humidification, raised-floor exposure, leak zones, Form-C dry contacts, and auxiliary alarm wiring context.
• Vertiv, Liebert Liqui-Tect LP6000 Leak Detection Installer/User GuideUsed for leak-detection cable layout, as-installed map updates, calibration, mapping and testing, BMS/NMS test coordination, alarm notification, zones, dry contacts, alarm history, reset, and fault context.

Related guides