How to Read a Sprinkler System for What It Was Designed to Do

Why this matters

Any trade can end up at a riser being asked a question about a sprinkler system: whether the building is covered for a new use, whether a ceiling change matters, what an insurer's letter means. You do not need to be a sprinkler designer to answer usefully, because the design intent of an installed system is recoverable from four physical artifacts already in the building. What you need is the discipline to say which of the four you read it from, and to hand back a named question with a named owner wherever they disagree or one is missing. A guess dressed as a finding is worse than no finding, because somebody builds on it.

This walk is read-only, deliberately: operating a valve, a test connection or a release changes the state of a system that spends its life doing nothing visible, and changing that state is a job with its own procedure and paperwork.

What you are trying to recover

Six fields. Write them as a form before you walk, one set per system, because a building routinely has three. System type: wet, dry, preaction with which interlock, or deluge, since response and demand follow from it. Design basis: hazard classification, density, area of operation, hose allowance. Head data: type and orientation, nominal K-factor, temperature band, response classification. Supply basis: where the water comes from and the date of the flow test the design rests on. Current state: valve positions, supervisory and trouble signals, open impairments. Occupancy match: whether what is in the space today is what the design assumed.

That last field most often produces a real finding and is the only one needing no fire protection expertise: you compare what is on the floor against what the design says.

Before you start: what you do not touch

• Do not close, open, exercise or "just check" any control valve. Closing one starts an impairment whether or not anyone files it, and the commonest way a system is found dead after a loss is a valve reopened without being confirmed fully open and secured.
• Do not open an inspector's test connection, a main drain or an auxiliary drain, which flow water and send a waterflow signal that gets logged as false. Do not operate a manual release station on a preaction or deluge system: it releases the valve, and the valve does not care why you pulled it. Do not break into a pressurized dry or preaction system, which holds stored energy in the air volume as well as water behind the clapper. Read the fire alarm control unit's display; do not open it, because the panel is energized and the display carries everything this walk needs.
• Read heads from the floor with optics. If a head must be reached, that is lift work with its own fall protection, not a reason to climb racking, which is not a ladder. Do not walk a rack aisle while lift trucks are operating: stop the aisle and confirm with the operators.

Station one: the riser

Start here, because the riser answers the system type and hands you the design basis on a plate.

1. Read the hydraulic design information sign. Every system installed under a modern adopted edition of NFPA 13 carries a permanent sign at the riser giving the area covered, the hazard classification or commodity class, the design density and area of operation, the hose allowance and the demand at the riser base. Photograph and transcribe it. Artifact one.
2. Read the valve arrangement. A wet system has a riser check or alarm check valve. A dry system has a dry pipe valve with an air gauge above and a water gauge below, plus an air supply and usually a quick-opening device. A preaction system has a preaction valve, a releasing panel or interface, and supervisory air. A deluge system has a deluge valve and no thermal elements downstream. The trim tells you which; the door label frequently does not. Artifact two.
3. Read both gauges and record them. On a dry or preaction system they are the whole condition in two numbers.
4. Record every control valve, its position, and whether it carries a supervisory switch and a tag. An untagged valve in a locked room is supervised by nothing but whoever last walked past.

Station two: the ceiling

1. Read three to five heads per protected area from the floor and log type, orientation, K-factor, temperature band and response classification. Markings carry manufacturer, model, nominal K-factor and temperature rating, and the band is colour coded. Artifact three. If any two heads in one area disagree without a reason you can name, that is a finding.
2. Sight for anything installed after the sprinkler system. New ductwork, a dropped grid, a light cove, a cable tray, a banner, a mezzanine. Anything between deck and deflector changes the geometry the design assumed, and none of it appears in any record.
3. Look at what is on the floor and how high it is stacked. Storage height, packaging, racking, clearance from top of storage to deflector. The occupancy match field, answerable without any other artifact.
4. Check the spare head cabinet against what you logged overhead. Spare stock that does not match installed heads is the cheapest evidence heads changed without records.

Station three: the records and the panel

1. Read the fire alarm control unit display and record every alarm, supervisory and trouble condition showing now, plus the event history if the panel offers one. A supervisory signal means a protective feature changed state, which on a sprinkler system usually means a control valve is not where it should be. Repeated low-air events mean a dry or preaction pipe leaking toward an unintended trip.
2. Read the impairment log. For each open impairment note the coordinator, the compensating measure and the expected close-out. An impairment with no compensating measure is an outage with a form attached.
3. Read the most recent reports under NFPA 25, in the edition the authority having jurisdiction has adopted and amended, and under NFPA 72 on the alarm side. Artifact four. Note what was tested and what was not.

Which artifact wins

The sign proves what the system was designed to, not that it still matches. The valve trim and the head markings prove what is installed right now and beat the sign on their own subjects, because they are the hardware. The reports prove only that the system was in its installed condition on the test date, because NFPA 25 inspection is written to confirm that and expressly not to evaluate whether the design suits the occupancy. So where the physical evidence and the paper disagree, the physical evidence wins on what exists and the paper wins on what was intended. The gap is the finding. It is not yours to close, and the correct output is a written question routed to a named owner: the designer for demand or classification, the owner for occupancy, the authority having jurisdiction for acceptance, the inspection, testing and maintenance contractor for condition. That authority is a named role with power to interpret and enforce, commonly a fire marshal or building official, not a synonym for whoever inspects. Write each finding as three columns, what you observed, which artifact it came from, and who owns the answer, never a recommendation no artifact you named supports.

The worked walk: one building, four systems, three findings

A two-story building: ground floor retail on a wet system, an exterior dock canopy on a small dry system, second floor offices on wet, a ground-floor server room on preaction.

Second floor offices, wet system. The sign reads light hazard, density 0.10 gpm per sq ft over an area of operation of 1,500 sq ft, hose allowance 100 gpm. Sprinkler demand is 0.10 times 1,500, or 150 gpm, plus the 100 gpm allowance for 250 gpm total at the point of connection. Valve trim confirms wet. Heads logged from the floor: concealed pendent, K-5.6, ordinary band. Four in a recently partitioned corner read standard response where every other head on the floor reads quick.

Finding one. Most adopted editions of NFPA 13 require quick-response heads throughout light hazard occupancies, so those four are both a deviation and a mismatch with their neighbours. Route: the sprinkler contractor, replacing them matched on type, K-factor, band, response class and trim together, plus the authority having jurisdiction if the partition work's permit status is in question.

Dock canopy, dry system. Both gauges within the range marked on the trim. The panel's event history shows six low-air supervisory events in 90 days, all acknowledged, none investigated. Two auxiliary drains on the pitched main, neither tagged, no seasonal drain entry in the record.

Finding two. Six low-air events in 90 days is an air supply chasing a leak, and a leaking dry system is one unattended weekend from an unintended trip. Undrained auxiliary drains are separately how a trapped section freezes solid with nothing at the panel to say so. Route: the inspection, testing and maintenance contractor, drains onto a seasonal schedule. Nothing was drained during this walk.

Server room, preaction system. Valve trim and releasing arrangement read double interlock: both a detection event and a loss of air are required. The releasing detection zone reads normal. The sign reads ordinary hazard group 1, density 0.15 gpm per sq ft over 1,500 sq ft.

Finding three. The adopted edition applies an area of operation increase to double interlock preaction systems, commonly 30 percent at the same density, and the sibling wet versus dry card owns that rule. Applied here the area re-bases from 1,500 to 1,950 sq ft and the demand from 225 to 292.5 gpm at the same 0.15 density. The sign shows 1,500 sq ft. Either it is stale, the design predates the requirement, or the increase was not applied. Route: the designer to confirm which, and the authority having jurisdiction on acceptance. Not a field determination, and not resolved by reading the sign harder.

Corrections, printed.

• Raw area of operation, server room: 1,500 sq ft from the sign. Correction: the double interlock increase re-bases the area rather than adding to the density. Corrected area 1,950 sq ft, demand 292.5 gpm, a difference of 67.5 gpm.
• Raw hose allowance, offices: 100 gpm, added at the point of connection on top of the sprinkler demand, not folded into the density arithmetic. No correction, stated so it is not double counted.
• Raw head log, offices: heads match on K-factor and temperature band. Correction: the five fields are gates, not a score, so matching four is still a fail. Corrected reading: rejected on response class.
• Raw inspection report: passed. Correction: an NFPA 25 report confirms installed condition, not design suitability. Corrected reading: consistent with all three findings, contradicting none of them.

Sibling-rule check on this walk. Demand was computed as density times area of operation throughout, never from a head count, consistent with the coverage card. The double interlock area increase came from the wet versus dry card rather than being re-derived. Head substitution was specified as matching all five fields together, consistent with the head types card, and response class and band were treated as two independent numbers, consistent with the card on what a head is waiting for. No valve was closed, no drain or test connection opened, no release operated, no impairment created.

What a bad version looks like. It ends with a verbal answer at the dock: "yeah, you're covered." Every finding above was invisible to a look around: the standard-response heads are indistinguishable from their neighbours at a glance, the low-air events exist only in a panel history nobody scrolled, and the interlock question exists only in arithmetic between two artifacts nobody compared.

References

• NFPA 13, in the edition adopted and amended by the authority having jurisdiction, for the hydraulic design information sign, the light hazard quick-response requirement and the design area increase for double interlock preaction systems
• NFPA 25, in the adopted edition, for what a periodic inspection confirms, spare stock and impairment
• NFPA 72, in the adopted edition, for the panel's alarm, supervisory and trouble signals and alarm impairment
• See related: What a Wet Pipe and a Dry Pipe System Each Solve; Why Sprinkler Coverage Is a Density Problem, Not a Head Count