line tightness testing: how often pressurized and suction piping each need one

Ask how often underground piping needs a line tightness test and you get two right answers, because the federal rule splits on how the fuel moves through the pipe. Pressurized piping carries an automatic line leak detector plus an annual tightness test. Suction piping gets a tightness test at least every three years, and some suction piping needs no release detection at all. The intervals are set in 40 CFR 280.41 and the test itself is defined in 280.44.

What the test has to prove

Before the intervals, the specification. 40 CFR 280.44(b) says a periodic test of piping may be conducted only if it can detect a leak rate of 0.1 gallons per hour at one and one half times the operating pressure. One tenth of a gallon an hour, on a line held at fifty percent above the pressure it normally runs at.

The number is the whole standard. A method that cannot resolve 0.1 gph at that pressure does not meet the rule, whatever it is marketed as. It is also worth knowing what the number is not: it is a test of the pipe, and it says nothing about the tank, the sumps, or the dispensers.

Pressurized piping: two requirements, two clocks

For piping installed on or before April 11, 2016, 280.41(b)(1)(i) asks pressurized piping to do two things at once.

• An automatic line leak detector, meeting 280.44(a). The device has to catch a 3 gallon per hour leak at 10 psi line pressure within one hour, by restricting flow, shutting it off, or raising an alarm. Its operation gets tested annually under 280.40(a)(3).
• An annual line tightness test under 280.44(b), or monthly monitoring under 280.44(c) in place of it.

The two get conflated constantly, and the confusion is understandable, since both have the word leak somewhere near them and both come round once a year. They measure different things at different resolutions. The leak detector is a device sitting on the line permanently, catching a 3 gph loss, which is a leak you could nearly hear. The tightness test is a precise measurement taken once, resolving thirty times finer. A site that tested its leak detector in March and thinks the piping is covered for the year has done half of what 280.41(b)(1)(i) asks.

Monthly monitoring is the alternative to the annual test, and for most sites that means interstitial monitoring on double-walled piping with a sensor in the containment sump. If the monitoring is running and documented every 30 days, the annual tightness test comes off the calendar. If it lapses, the requirement does not go away, it reverts.

Suction piping: three years, or nothing

280.41(b)(1)(ii) gives suction piping a longer interval. A line tightness test at least every three years, or a monthly monitoring method instead.

Then it gives back the whole requirement to piping built a particular way. No release detection is required for suction piping that meets all five of these:

• The below-grade piping operates at less than atmospheric pressure.
• The below-grade piping is sloped so the contents drain back into the storage tank if the suction is released.
• Only one check valve is included in each suction line.
• The check valve is located directly below and as close as practical to the suction pump.
• A method is provided that allows the middle three to be readily determined.

The trade calls this safe suction, and the logic is physical. Below atmospheric pressure, a hole in the line draws air in rather than pushing product out, and a sloped line with a single check valve at the pump drains back to the tank when the pump stops. There is no head of fuel sitting in the pipe waiting to find the hole.

Read the fifth condition again, because it is the one that decides arguments. The design has to be demonstrable. An inspector standing at your site is entitled to ask how you know the line is sloped correctly and that there is exactly one check valve in the right place. A drawing, an installation record, or a photograph of the pump with the valve visible answers that. A verbal assurance from whoever installed it in 1998 does not, and a site that cannot show it is a site with a three year test it has not been doing.

Piping installed after April 11, 2016

Newer piping runs on a different clause. Under 280.41(b)(2), piping installed or replaced after April 11, 2016 gets monitored for releases at least every 30 days under 280.43(g), which is interstitial monitoring on secondary containment. Pressurized piping still carries an automatic line leak detector on top of that.

The safe suction exemption survives the change. Suction piping meeting those same five conditions still needs no release detection.

So the install date matters. A canopy replacement that swapped the piping in 2019 moved that run onto the 30-day rule even though the tanks underneath it did not move. Sites that have been through a rebuild often have two vintages of piping in the ground on two different requirements, and the compliance record has to say which is which.

Your state may want more

The federal rule is a floor. A state with program approval can set intervals above it, and several have.

Connecticut publishes the clearest example, a periodic testing table on the DEEP site that reads across every tank and piping category. On piping it sets the line tightness test for pressurized piping at annual under RCSA 22a-449(d)-104(b)(4)(C), the automatic line leak detector test at annual under 104(b)(4)(B), and the suction line tightness test at three years under 104(b)(5). Two piping categories move to a six month interval on both the detector test and the tightness test. The table carries a note saying a line tightness test is not required for safe suction systems, which tracks the federal exemption.

Connecticut also puts the piping interstitial space test and the liquid tight test on sumps and spill buckets at three years, under 103(d)(2) and 103(e)(2). Those are separate requirements from the line test and they land on their own clocks.

Read the table your own regulator publishes before you write a date on a calendar from the federal rule. Two operators fifty miles apart on opposite sides of a state line can owe genuinely different things on identical equipment.

What the record has to show

A test that happened and cannot be evidenced is a test that did not happen, as far as an inspection goes. For each line test, keep the date, which piping run it covered, the method used, the leak rate the equipment is rated to detect, the result, and who performed it. Keep the leak detector's annual operation test as a separate record with its own date, because it is a separate requirement.

Where you are relying on safe suction rather than testing, keep the evidence of the design instead. Assembling that file after an inspector asks is the expensive version.

How FastDragon handles it

FastDragon Compliance works the interval out from what the site actually is. Record the piping as pressurized and the calendar raises an annual line tightness test and a separate annual leak detector test. Record it as suction and the tightness test lands on a three year cycle. Record it as safe suction and neither one is raised, with the reason on the record so the next person can see why the site is quiet.

The next date comes off the day the last test was signed off rather than a fixed month, so a test done in June sets the following June. Marking one done asks for the report and files it against the requirement, which is what turns a pile of PDFs into an answer to a question. For sites in a state that asks for more than the federal floor, the state's own interval is what the calendar uses.

None of that is a substitute for reading the rule that applies to you. It is a way to stop the reading being something you do once and then rely on memory for.

Common questions

See the compliance calendar →