EN 12845 Annex K — 25-Year Sprinkler Inspection

Have twenty-five years of steel, copper, bronze, solder alloy, glass and spring proved the sprinkler's durability — or is it silently failing through corrosion, fatigue and scale? The Annex K sampling procedure answers that question.

A paper mill sprinkler system commissioned in 1995 went through Annex K testing in 2020. Six samples were sent to an accredited laboratory: the K-factor had dropped by 12 % on four heads, and RTI had risen from 92 to 178 through solder oxidation. One sprinkler still had not opened at 220 °C. The result was full replacement — 1840 heads over a six-month programme. The site's own view had been that "everything looked fine". A sprinkler head looks fine from the outside and can be finished on the inside; Annex K is the remedy for that illusion.

Annex K — the 25-year inspection procedure

Selecting samples

• A separate sample group for each sprinkler type (pendent, upright, sidewall)
• A separate group for each temperature rating (68, 79, 93, 141 °C)
• Minimum four sprinklers per group
• On systems above 5000 heads, one extra sample per 1000 sprinklers
• Samples taken from the oldest branches and the worst environment — humidity, corrosion, kitchens
• Samples marked, photographed and recorded with dates

Accredited laboratory testing

Samples are tested to the EN 12259-1 sprinkler product standard, covering:

• K-factor, the flow-pressure relationship
• RTI, the Response Time Index measuring thermal sensitivity
• Operating temperature — the actual trigger point
• Water distribution pattern
• Mechanical strength: spring load and contraction
• Analysis of corrosion residues

Acceptance criteria

• K-factor must not deviate by more than 5 % from the design value
• RTI must not rise more than 25 % above the manufacturer's specification
• Operating temperature must trigger within ± 5 % of nominal
• No sample may show a mechanical defect

If a sample fails

One failure means retesting double the number of samples of the same type. If that round also produces a failure, all similar sprinklers are replaced.

Sprinkler type and test interval

Sprinkler type / environment	First inspection	Repeat interval
Standard pendent/upright, dry environment	25 years	10 years
Fast response (RTI ≤ 50)	20 years	10 years
Dry pendent (liquid filled)	10 years	10 years
Corrosive environment (coastal, chemical, plating)	5 years	5 years
High temperature (≥ 141 °C)	10 years	5 years
ESFR / CMSA	20 years	10 years

Field error — samples taken from the wrong place

A shopping centre carried out Annex K testing after 26 years. Four samples were taken from the office floor ceiling — a clean, dry environment. They passed. Six months later a head on the restaurant floor opened spontaneously, its scaled spring having failed under pressure. The sampling had been done in the wrong place. Annex K requires samples from the worst environment, and the area under kitchen grease vapour and constant humidity had been skipped. A second sampling round showed that every sprinkler on the restaurant floor needed replacing. Sample selection strategy is at least as critical as the test itself.

Preparing for the test

• A physical system map, with sprinklers labelled by type, age and environment
• A candidate list — twice as many candidates as samples, for each type and risk environment
• Authority and insurer notified
• Sequential system shutdown under the impairment procedures
• Spare sprinkler stock, for the heads removed and any that fail
• A laboratory shipping plan covering type approval, packaging and transport
• Test results retained for the life of the system

Cost and duration

Item	Typical cost (€)	Duration
Sample removal and replacement	200–500 per sample	1–2 days
Laboratory testing (K, RTI, strength)	400–800 per sample	2–6 weeks
Fire testing, where required	1500–3000 per group	4–8 weeks
Report and certificate	500–1500	1 week
Complete package, 1000-sprinkler system	3000–5000	2–3 months

Annex K testing versus early replacement

For a 1000-sprinkler system:

• Annex K testing on four to nine samples: €3000–5000, with the great majority of the system staying in place.
• Replacing every sprinkler: €25 000–40 000, plus labour and system downtime.

Testing costs 10–15 % of replacement, which is why the 25-year test is economically rational. If it fails, replacement was unavoidable anyway; if it passes, the system gains another ten years.

Comparison with NFPA 25

NFPA 25 sets its sampling intervals as:

• Standard sprinklers: 50 years, then every 10 years
• Fast response: 20 years, then every 10 years
• Dry sprinklers: 10 years
• High-temperature environments: every 5 years
• Corrosive environments: more frequently, per the manufacturer's recommendation

NFPA allows 50 years for standard sprinklers, where EN 12845 is more conservative at 25. Fast response is 20 years in both. Laboratory and sample count requirements are comparable, and where the two are applied together the EN 12845 25-year interval satisfies both.

Link to Turkish regulation

BYKHY refers periodic inspection and maintenance to the Turkish standards, so the TS EN 12845 Annex K procedure applies in Turkey. Systems older than 25 years are still uncommon locally — most are 5 to 15 years old — but some industrial facilities built in the 1990s now need testing. Without a test certificate at year 25 an insurer will treat the system as beyond its service life and require either a premium increase or replacement. Accredited sprinkler test laboratories are limited in Turkey, so samples are generally sent to established European or US testing bodies.

Quick check list

• Is the system approaching 25 years — or 10 for dry, 5 for corrosive environments?
• Sampling plan covers type, temperature rating and worst environment.
• An accredited laboratory has been appointed.
• Spare sprinkler stock ready.
• Impairment measures planned for the shutdown.
• Test results archived for the life of the system.
• Insurer and authority notified.
• The next test is already in the calendar, ten years out.

Frequently Asked Questions

How many samples does Annex K require?

At least four per sprinkler type and temperature rating, plus one extra per 1000 sprinklers on systems above 5000 heads — taken from the oldest branches in the worst environment.

What are the acceptance criteria?

K-factor within 5 % of design, RTI no more than 25 % above the manufacturer's figure, operating temperature within ± 5 % of nominal, and no mechanical defects.

Is testing cheaper than replacement?

Substantially. Testing a 1000-sprinkler system costs roughly 10–15 % of replacing every head, and a pass buys another ten years of service.

Why does sample location matter so much?

Because a clean office ceiling tells you nothing about heads under kitchen grease vapour. Annex K explicitly requires sampling from the worst environment in the building.

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 — 86+ Engineering Calculators

MEP Calc bundles 86+ 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