What IGEM/UP/1B Edition 4 means for gas engineers from 1 October 2026

Every gas engineer knows the tightness test. Isolate, pressurise, watch the gauge, note the drop, compare it against a limit you probably learned early enough that you no longer think about it. It is one of the most repeated actions in the trade. From the 1st of October, the limit you are comparing against changes, and so does what happens when the needle moves more than you would like.

IGEM/UP/1B Edition 4 was published back in March 2026, and for the last six months it has been running alongside the old Edition 3, giving engineers and businesses time to move over. That grace period ends on 30 September. From 1 October, Edition 3 is withdrawn and Edition 4 is the only version Gas Safe registered engineers can work to. This is not a paperwork update sitting in the background. It changes a test almost every registered engineer runs multiple times a week.

At a glance

• Edition 3 withdrawn: 30 September 2026
• Edition 4 mandatory: from 1 October 2026
• Applies to: small NG, LPG and LPG/Air installations, meter capacity up to 16 m³/h, max 2 bar at the ECV outlet
• Core change: permissible pressure drop is now based on calculated Installation Volume, not the meter's badged capacity
• New mandatory step: a pipework-only retest when there is any perceptible gauge movement within the old permissible limit

Purge volume calculation also changes to 1.5 x Installation Volume for every job.

Why meter size was always a slightly odd way to judge a leak

The old approach set your permissible pressure drop by the meter fitted, a U6, a U16, whatever the badge said. It made sense as a shortcut. It was also never quite measuring the right thing. A meter's rated capacity tells you how much gas it can pass, not how much gas is sitting in the pipework, fittings and meter at the moment you test it. Two installations with the exact same size leak can show two different pressure drops, purely because one has a larger installation volume than the other. The bigger system dilutes the same leak into a smaller reading. That is not a flaw you would notice on any individual job, but across the industry it meant the test was inconsistent in a way nobody could see from the gauge alone.

Installation Volume fixes that by making the calculation specific to the job in front of you. IV is now built from three parts: the pipework volume, the meter volume, and an allowance for fittings, worked out as ten per cent of the pipework figure. Add those together, and your permissible drop comes from a table keyed to that number and the gas type, not the meter's badge. It is more calculation than most engineers were doing before. It is also a fairer test.

The bit that will actually slow you down: the pipework-only retest

This is the change most engineers will feel on the tools, not on paper. Under Edition 4, if your initial test shows any perceptible movement, even movement that is still inside the permissible limit, you cannot simply write it up and move on. You now have to isolate every appliance from the supply and retest the pipework on its own. On that second test, there is no allowance at all. Perceptible movement is now defined in numbers, 0.25 mbar on a water gauge, 0.2 mbar on a high resolution electronic gauge, and any movement above that is a fail.

The logic is sound. It separates a leak that is genuinely in the fixed pipework from one sitting in an appliance connection, which is exactly the distinction you need before you start taking things apart. The cost is time. A test that used to end with a note in the log now sometimes ends with a second test, and if that second test fails, with a fault you have to trace and repair, or make safe, on the spot. There is also no going back to Edition 3's more forgiving criteria to rescue a job that fails under the new rules. If it fails under Edition 4, it stays failed until it is fixed.

Where this leaves engineers on 2 October

The honest way to describe this update is that it trades a bit of speed for a lot more certainty. The benefit is real. A test that scales to the actual installation, with a numerically defined pass and fail rather than a judgement call, protects engineers as much as it protects customers. Nobody wants to be the person who signed off a borderline result that turns out to matter later. The pipework-only retest gives you a useful diagnostic step, not just an extra hoop.

The friction is just as real. Engineers now need to calculate IV correctly on every job, using pipe diameter and length rather than a number they already know from the meter badge, and a wrong calculation undermines the whole test. Jobs that would have passed cleanly under Edition 3 will now sometimes need a second test and, occasionally, a return visit. On a day already full of jobs, that adds up.

This is exactly the kind of change where the certificate and the job sheet matter as much as the test itself. Which standard was applied, what the calculated IV was, what the readings were on both tests, all of that needs to be recorded properly, not remembered. Businesses running that through a proper job management system, like Joblogic, rather than a paper certificate and a good memory, are the ones who will be able to show their working if a job is ever questioned months down the line. That is not a reason to buy anything. It is a reason to make sure whatever system you already use is actually capturing it.

From 1 October, this is simply how the test is done. The engineers who are already comfortable with the calculation and the retest step will barely notice the deadline pass. The ones who leave it until the last week of September are the ones who will find out the hard way that the old shortcut no longer applies.