FM Global DS 2-81: ITM Frequency Table and Common Findings

The way to read DS 2-81 correctly is to treat it not as a calendar but as a set of tables. Section 2.1 (p.4) sets the frame: Tables 1-10 contain both frequency-based and event-driven activities, and the frequency-based ones are listed with a baseline frequency.

The source is FM Global DS 2-81, Fire Protection System Inspection, Testing and Maintenance (April 2019, Interim Revision October 2025, 74 pages). For a general walk-through of the activities see the DS 2-81 ITM field guide, and for the NFPA side see sprinkler ITM inspection frequencies.

The Programme Framework

Section 2.2.1 (p.4) wants ITM done by trained personnel or qualified contractors; Section 2.2.1.1 asks for initial and annual refresher training and a competent back-up group to step in if the primary personnel become unavailable. Section 2.2.2 gives the minimum content of the records: the systems and equipment covered, the type of ITM, the results and any corrective actions needed; records are kept for at least one year. Section 2.2.3 asks for the programme itself to be audited at least annually, and the audit includes the timeliness of ITM work-order completion.

Sections 2.3.1 and 2.3.5 (p.5) tie impairment management in from two sides: when protection is taken out of service for ITM, an impairment programme (FM Red Tag Permit System or equivalent) is used; and if ITM reveals that a system is already out of service, the same programme applies. Inoperable components, poor performance and poor physical condition are listed as instances where a system may be considered out of service.

Control Valves: Table 1

Section 2.4.2.3 (p.6) defines "secured": a valve is secured when its operator cannot be moved more than one turn toward closed (or not at all for quarter-turn valves). Each valve is secured separately with its own lock and chain, combination locks are not used, and seals or breakaway locks are accepted only on valves of 1.5 in. (38 mm) nominal diameter or smaller, or controlling five or fewer sprinklers.

Section 3.1.2.5 (p.41) states the limit of supervisory alarms: they do not prevent tampering, they only detect and report it, typically within 2-3 turns toward the closed position. For what this looks like on site, see closed valve failures.

Activities Common to All Sprinkler Systems

Table 2a (Section 2.5.1.1, p.7-8) covers every system type. The key items:

• Quarterly: test waterflow alarms by flowing water through the system test connection (annually for antifreeze systems). The table notes that receipt of the alarm should take less than 60 seconds per DS 2-0.
• Annually: test building and process interlocks actuated by waterflow alarms; check the water supply upstream of each riser for significant obstruction with a main-drain flow test (ideally after the annual valve exercise); inspect sprinklers, nozzles, piping, pipe supports and seismic protection for damage and poor condition; physically inspect at least 10% of sealed concealed sprinklers per room and visually inspect the rest.
• Every 3 years: sample test sprinklers with fusible elements rated 360 °F (180 °C) or more that are exposed for long periods to 300 °F (150 °C) or higher.
• Every 5 and 15 years: sample test recalled O-ring sprinklers and dry-type (dry pendent) sprinklers.
• Event-driven: replace dry-type sprinklers manufactured before 2003 when found; after a fire, replace non-operated sprinklers within at least 20 ft (6 m) of any operated sprinkler.

The concealed sprinkler item is the one most often skipped on site: it is physically checked that the cover gasket has not adhered to the ceiling, each year sprinklers not selected in earlier years are chosen, and if a deficiency is found in one sprinkler, every sprinkler in that room is physically inspected.

Section 2.5.1.3.1 (p.8-9) lists the damage to look for: green discoloration or white deposits showing leakage from the orifice button and seal, surface corrosion, residue and dust accumulation, paint, exposure to temperatures within 50 °F (28 °C) of the sprinkler's rating, signs of freeze damage, and mechanical impact on the deflector.

The extra activities for wet systems in Table 2b (p.10) are short: every 5 years, check systems fed by an open water supply for obstructive debris regardless of pipe material; every 5 years in areas known for hard water, check sprinkler-pipe connections for mineral deposits; annually, test the solution in antifreeze systems.

Dry, Preaction and Deluge Systems: Table 2c

Section 2.5.3.3 (p.10) sets which items apply to which system: items 1-16 for dry systems, 1-17 for preaction and vacuum systems, 1-16 and 18-20 for refrigerated-area systems, and 1-16 and 21-24 for deluge and fixed water spray systems.

The effect of nitrogen shows numerically in this table: the full-flow trip test interval goes from 3 years to 10, and the periodic obstruction check excludes systems originally installed with nitrogen (item 15). For nitrogen and corrosion see the DS 2-1 corrosion article.

Why the air leakage rate is measured annually is also written in the table (item 10): to identify systems prone to false trips during power outages, to decide when action is needed to reduce leakage, and, for pilot lines, to make sure loss of air supply will not trip the system for the design duration of the water supply or 2 hours, whichever is greater.

Obstruction: the 120 ml Threshold

Section 2.5.1.2.1 (p.8) sets the scope of an investigation: the feed main, at least one cross main and at least three branch lines. The methods are a flushing investigation, a videoscope inspection or an ultrasonic localized guided wave evaluation.

Section 2.5.1.2.4 gives the threshold; the system is treated as obstructed if any of these is present: approximately 1/2 cup (120 ml) or more of debris in a cross main; debris pieces in the piping large enough to plug a sprinkler orifice; obstructed flow from a branch line; videoscope or ultrasonic results showing the system is obstructed. Section 2.5.1.2.6 adds a separate requirement: if deposits (tubercles) are found attached to the internal pipe walls, they are physically removed or the pipe section is replaced, and a solution is developed with reference to DS 2-1 to suppress the corrosion mechanism.

On flushing, Section 3.1.3.3 (p.45) says a velocity of at least 10 ft/s (3 m/s) is needed to scour the pipe and lift foreign material to an aboveground outlet. Table 11 gives the flows: DN100, 390 gpm (1,475 L/min); DN150, 880 gpm (3,325 L/min); DN200, 1,560 gpm (5,895 L/min); DN250, 2,440 gpm (9,225 L/min); DN300, 3,520 gpm (13,305 L/min). The same section adds that if large material is suspected, a larger outlet will be needed to pass it and to create the 750 gpm (2,839 L/min) flow needed to move it.

Common Findings and Root Causes

Obstruction sources (Section 3.1.3.1, p.42). The data sheet says it plainly: dry-pipe systems are involved in the majority of obstructed sprinkler systems, and pipe scale is the most frequent obstructing material. Systems kept alternately wet and dry over the years are particularly prone to scale accumulation; in systems that stay dry, condensation of moisture in the air supply can form a hard scale along the bottom of the pipe. When sprinklers open, this scale breaks loose, travels along the pipe and plugs sprinklers or fittings.

Careless installation and repair. The same section lists obstructing materials found in the field: wood, paint brushes, buckets, gravel, sand and gloves. In welded systems and systems with cut holes for quick-connect fittings, the cutout disks (coupons) may have been left inside the pipe.

Raw water sources. Through poorly arranged or inadequately screened intakes, fine compacted materials such as rust, mud and sand, and coarse materials such as stones, cinders, cast-iron tubercles and wood chips, can be drawn into the system from the bottom of rivers, ponds or open reservoirs.

Calcium carbonate deposits. Item E of Section 3.1.3.1 (p.42-43) is a particular warning: this type of sprinkler obstruction cannot be detected or corrected by normal flushing procedures; it can only be found by removing and inspecting sprinklers in suspected areas. The most likely locations are systems not connected to public water but supplied untreated, directly from wells or surface water, in very hard water areas.

Valve problems (Section 3.1.2.6, p.42). Indicator posts can become inoperative from corrosion or freezing due to a leaking valve, or be broken by frost action or by vehicles. Indicator targets can be misadjusted and prevent full valve travel, or even be set to read OPEN while the valve is closed. A valve gate can separate from the operating stem through corrosion or excessive strain when forced against obstruction, heavy deposits or friction.

Field Note

More than the numerical frequencies of DS 2-81, two sentences change the work on site. The first is in Section 1.1 (p.3): if the pump is isolated by closing a valve during churn testing, the ITM test is still completed, but the total health and connection of the system is not evaluated and understood. The second is in Table 7 item 2: do not isolate the pump during churn testing. On site, pump tests are often run with a valve throttled so that "it isn't noisy and the pressure doesn't rise"; the form gets filled in, but no information is produced.

The second lesson is on the documentation side. Section 2.2.2 asks for records to be kept for at least one year, but the logic of Sections 2.5.1.2.3 and 2.5.2.2.2 requires more: different portions of the system examined in each investigation, and the examined portions documented. That cycle cannot be built on a single year of records. In practice, keeping an "examined sections map" on the facility's sprinkler plan is the one step that turns DS 2-81 from a calendar into a management tool.

Checklist

• ☐ Initial and annual refresher training for ITM personnel recorded; back-up group defined (Section 2.2.1.1)
• ☐ ITM records include system, ITM type, result and corrective action fields (Section 2.2.2)
• ☐ Control valves on weekly visual, monthly physical and annual full-travel exercise (Table 1)
• ☐ Valves secured individually with lock and chain; no combination locks (Section 2.4.2.3)
• ☐ Waterflow alarm test quarterly; alarm received in under 60 seconds
• ☐ At least 10% of concealed sprinklers per room physically inspected
• ☐ Annual partial-flow trip test compared with the last full-flow test (Table 2c item 13)
• ☐ Full-flow trip test every 3 years; every 10 years on nitrogen systems (item 14)
• ☐ Air leakage rate measured annually and trended (item 10)
• ☐ Obstruction investigation covers the feed main, one cross main and three branch lines
• ☐ The 120 ml threshold and the other three criteria are written on the evaluation form (Section 2.5.1.2.4)
• ☐ Examined pipe sections mapped on the plan, with different sections chosen each time
• ☐ The pump is not isolated during churn testing (Table 7 item 2)

Frequently Asked Questions

How often are control valves checked under FM DS 2-81?

Table 1 (Section 2.4.2.1, p.5-6) defines four activities. Control valves are visually inspected weekly for full-open, secured and accessible conditions; for enhanced security indicating valves this inspection is semiannual. Control valves, including PIV, WPIV, NRS, curb-box and non-indicating butterfly valves, are physically tested monthly for the full-open position. Supervisory (tamper) alarms are tested semiannually. All control valves get an annual full-travel exercise, recording the number of turns to close and to reopen.

How often is a full-flow trip test done on a dry system?

Table 2c item 14 (p.12) asks for a full-flow trip test every 3 years, or every 10 years for systems with nitrogen. A videoscope inspection or an ultrasonic localized guided wave evaluation can be used instead. The criterion is that water reaches the hydraulically remote area within the specified time; unless an FM data sheet for the protected occupancy states otherwise, the maximum water delivery time is 60 seconds. In addition, item 13 requires an annual partial-flow trip test.

What threshold makes a system count as obstructed?

Section 2.5.1.2.4 (p.8) treats the system as obstructed if any of four conditions is present: approximately 1/2 cup (120 ml) or more of debris in a cross main; debris pieces large enough to plug a sprinkler orifice; obstructed flow from a branch line; or videoscope or ultrasonic evaluation results showing the system is obstructed. An obstructed system is treated as impaired protection until the piping is completely flushed.

How is the churn test of a diesel fire pump done?

Under DS 2-81 Table 7 item 2 and Section 2.9.3.1.2, a diesel pump is started weekly in automatic mode via pressure drop and run at churn for at least 30 minutes (electric pumps monthly, at least 10 minutes). The pump is not isolated by closing a valve during the test; suction and discharge pressures, RPM, oil pressure and coolant temperature are recorded. Since the October 2025 interim revision, this weekly test can also be done with an FM Approved automated fire pump test system without personnel present (2.9.3.2).

References

1. FM Global Property Loss Prevention Data Sheet 2-81, Fire Protection System Inspection, Testing and Maintenance, April 2019, Interim Revision October 2025. Sections used: 1.1 (p.3), 2.1-2.3 (p.4-5), 2.4.2 and Table 1 (p.5-6), 2.5.1 and Table 2a (p.6-8), 2.5.1.2-2.5.1.3 (p.8-9), Table 2b (p.10), 2.5.3 and Table 2c (p.10-13), 2.9.1.1, Table 7 and 2.9.2-2.9.3 (p.16-19), 3.1.2.5-3.1.2.6 (p.41-42), 3.1.3.1-3.1.3.3 and Table 11 (p.42-45).
2. FM Global data sheet list: https://www.fmglobal.com/research-and-resources/fm-global-data-sheets
3. General walk-through of DS 2-81 activities: https://selvi.org/en/blog/fmglobal/ds-2-81-inspection-testing-maintenance
4. Inspection frequencies on the NFPA side: https://selvi.org/en/blog/nfpa/sprinkler-itm-inspection-frequencies
5. Corrosion mechanisms and nitrogen: https://selvi.org/en/blog/fmglobal/ds-2-1-sprinkler-corrosion
6. Closed valve failures: https://selvi.org/en/blog/saha/closed-valve-failures

SprinkCalc — Fire Sprinkler Design Across Three Standards

SprinkCalc covers hazard classification, design density and area, K-factor selection, water demand and hydraulic calculations for NFPA 13, FM Global and BS EN 12845 in a single iOS app, and exports a professional PDF report.

Download SprinkCalc on the App Store

MEP Calc — 110+ Engineering Calculators

MEP Calc bundles 110+ engineering modules in one iOS app: 21 fire calculations plus heating, cooling, HVAC, plumbing, steam and natural gas.

Download MEP Calc on the App Store