A pressurised line pushes product out of any hole it has, at a rate set by the pump rather than by gravity. The automatic line leak detector is the device that catches the fast version of that, and 40 CFR 280.44(a) gives it one number to hit. Three gallons per hour, at ten pounds per square inch, within one hour.
The one number
280.44(a) allows methods that alert the operator by restricting or shutting off the flow of product through the piping, or by triggering an audible or visual alarm, and only if they detect leaks of 3 gallons per hour at 10 psi line pressure within 1 hour.
Three gallons an hour is a large leak. It is roughly a gallon every twenty minutes going into the ground, and it is meant to be: the ALLD exists to catch a catastrophic line failure fast, on a system that is actively pumping product under pressure. Fine resolution belongs to the other half of the requirement.
Any of the three responses satisfies the rule. Most modern detectors restrict flow, which is why a failing line often announces itself as a dispenser that has gone slow rather than as an alarm. A pump that suddenly delivers at a trickle is a symptom worth treating as a line problem until proven otherwise.
Half of a two-part requirement
For piping installed on or before April 11, 2016, 280.41(b)(1)(i) asks pressurized piping for two things: an automatic line leak detector, and either an annual line tightness test or monthly monitoring.
The detector never satisfies the second half. This is the most common misreading in the whole of piping release detection, helped along by both requirements landing annually and both containing the word leak. One is a device that watches the line permanently at a coarse threshold. The other is a precise measurement, resolving 0.1 gallons per hour, taken once.
For piping installed or replaced after April 11, 2016, 280.41(b)(2)(i) pairs the detector with 30-day monitoring under 280.43(g) instead.
Testing it means simulating a leak
The annual test comes from 280.40(a)(3)(iii), inside the general operability testing requirement, and its wording is worth quoting: test operation to meet criteria in 280.44(a) by simulating a leak.
That makes the test active. A technician introduces a controlled loss and confirms the device responds inside the threshold. Opening the sump, looking at the detector and noting it appears undamaged is an inspection, and it is not what the rule asks for.
It also means the test produces a result worth reading. A detector that responds at four gallons an hour instead of three has failed and needs replacing, and that is a finding you only get from a simulated leak.
Where it sits in the year
The ALLD test is one line item in the annual operability test that covers the console, the probes and sensors, and the vacuum pumps and pressure gauges as well. One contractor visit usually closes all of them.
Keeping them as separate records still matters, because the components fail separately and the report has to say which one was found wanting. 280.45(b)(1) requires the annual operation test record to list each component tested and indicate whether each meets the criteria or needs action taken. A single line saying the annual test passed does not meet that.
Suction lines do not have one
The detector requirement is written for piping conveying product under pressure. Suction piping is covered separately, and safe suction designs are exempt from release detection entirely.
Worth stating because a site with both configurations, which is common where a canopy was rebuilt around older infrastructure, has different obligations on different runs of pipe on the same forecourt.
How FastDragon handles it
The detector test appears on the calendar only where the site profile records pressurized piping, and it sits as its own item rather than folded into the line tightness test. Two requirements, two records, two dates.
The interval follows the state. A Connecticut site takes the annual figure from the DEEP periodic testing table, and where a piping category is on the six month cycle, that is the number the calendar uses rather than the federal annual.
Common questions
What performance does an ALLD have to meet?
40 CFR 280.44(a) sets one threshold: it must detect a leak of 3 gallons per hour at 10 pounds per square inch line pressure within 1 hour. It meets that by restricting or shutting off flow through the piping, or by triggering an audible or visual alarm. Any of the three responses satisfies the rule.
How is the annual test actually performed?
By simulating a leak. 280.40(a)(3)(iii) says the test of operation must meet the criteria in 280.44(a) by simulating a leak, so the technician introduces a controlled loss and confirms the detector responds within the threshold. A visual inspection of the device is not the test.
Does an ALLD satisfy my piping release detection on its own?
No, for piping installed on or before April 11, 2016. 280.41(b)(1)(i) requires both an automatic line leak detector and either an annual line tightness test or monthly monitoring. The detector is one of two requirements, never both. For piping installed or replaced after that date, 280.41(b)(2)(i) requires 30-day monitoring plus the detector.
Do suction lines need one?
No. The automatic line leak detector requirement in 280.41(b)(1)(i) applies to piping that conveys product under pressure. Suction piping is covered by 280.41(b)(1)(ii), which asks for a three year line tightness test or monthly monitoring, and exempts safe suction designs entirely.
Does Connecticut use the annual interval?
Its periodic testing table sets the automatic line leak detector test at annual for most piping categories under RCSA 22a-449(d)-104(b)(4)(B), and moves two categories to a six month interval. Read the row for your own piping category.