Yes it would, but part of that is that above a set pressure they can start venting.
The idea is that an increase in pressure relative to ambient will cause a needle (low flow) or disc (high flow) valve to open or close, pressure in the pipe higher than the desired delta will cause the valve to close and vice-versa.
The venting is a safety measure to protect downstream equipment in case of regulator failure (for instance, if it seizes too far open, or the seals are porous and the consumers are all not using gas which would slowly cause the pressure in the house side to rise to the same pressure as the feeder line), and normally this would be to let little bits of gas escape, but if the line pressure were high enough it would allow a constant stream of gas to escape from regulators that were 'borderline' or seized up. It could also cause the membrane under the spring to break, and that would let even more gas out through the seal between the tap stem and the regulator housing, and any consumers that were 'on' at the time would have to deal with the overpressure directly which they would not be designed for. They can handle a bit of overpressure but not for ever and not beyond certain set limits. Especially aluminum burner housings would simply burn up after some time.
In some lines there are burst plates as the final safety, those really should never go but when they do it is definitely a spectacle if there is an ignition source nearby. A burst place is a purposefully weakened structure that is bolted (with a seal) on top of a gasline so that when the pressure increases above the burst plate's rating the plate will rupture. This does not self reset for obvious reasons and will need an on-site technician to repair and would come with a serious investigation into what went wrong upstream.
There is also the opposite device, an emergency shutdown valve that automatically closes when the flow gets either too high or goes negative. The former to ensure that a ruptured pipe will not vent enormous quantities of explosive gas in to the atmosphere, the latter to avoid having an air/gas mixture in the lines.
Gas infrastructure is quite interesting, and very expensive to install and maintain.
Up to a point, yes. But fundamentally a regulator consists of a pipe in and a pipe out with a valve in the middle controlled by the downstream pressure. A high enough pressure in the input pipe can rupture the valve seals (which are rubber). Then the full pressure goes into the downstream pipe (which isn't rated for high pressure) and it leaks.
The system should have safety overpressure relief valves, but obviously something went wrong.
> The system should have safety overpressure relief valves
It does, but those will vent to the atmosphere, if there is enough overpressure and it lasts for long enough the amount of gas released would be considerable.
The whole thing sounds as if that is what happened, an overpressure situation lasting longer than a few moments allowed weak spots in lines and homes to fail leading to fires and explosions.
What? How come the over-pressure safety equipment is not simply locks down the pipe, building backpressure to let the other side know about the problem? Or no such passive device is possible?
The over-pressure protection sits on the far side of the pressure reducer.
Imagine a tube that should be at 1.2 atm, and a feeder line that is at 5 atm. The equipment and the lines on the far side of the regulator may be able to deal with 3 but no more than that. If all the consumers are off and you do not bleed excess gas from the bypass leakage into the atmosphere the pressure on the far side will slowly increase until it is 5 atm, the pressure of the feeder line (no seal is perfect). So you need some way to get rid of that, and normally that is what the vent does, so it will allow miniscule amounts of gas to pass out in order to not let the far side go overpressure.
But if the feederline suddenly jumps to 10 atm, ruptures the seal or if the regulator sticks 'open' when that spike happens then the only thing that stops your internal line to go to that pressure is the bleeder (and in extreme cases, a burst plate that will break). This would allow a lot of gas to be released.
This is also why in some countries regulators are installed outside of the house rather than inside.
Another avenue for gas to escape if the seal breaks is the atmospheric port that allows the pressure of the chamber on the atmospheric side to be the same as ambient, in a sealed housing that would not be the case. Gas pressure regulation is not just about reduction, it is also to output a fixed delta relative to ambient otherwise your stove would make much larger flames if the atmospheric pressure was low!
Those regulators are works of art, totally passive and yet with a whole raft of features.
I am surprised that only some countries require the regulator to be outside. Just seems like a relatively simple thing to do. Does having the regulator outside require more maintenance or something?
You have to remember that natgas is extremely old, like mid industrial revolution era. Homes had natgas before they had electricity, usually.
In the early days they billed on the number of gas jets for illumination but once gas stoves and furnaces became a thing they needed some way to measure, and there's some weird technology out there involving something like an archimedes screw sitting in a couple inches of anti-freeze driving low friction clockwork. Now a days its probably optical sensors instead of clockwork but whatever.
Two engineering topics are being missed; flow rate thru an orifice is strongly non-linear, and there are no hermetically sealed large scale transport technologies. Its staggeringly impressive that like 99% of electricity pumped into the grid makes it out the other end. Its not practical from an engineering standpoint to operate a leak free material transport pipeline. There's plenty of engineering guidelines such that a leak rate ten times lower than could theoretically cause an explosion is quite acceptable, until some idiot accidentally applies a 10x overpressure forcing open orifices permanently and increasing the leakage flow rate 200x and then kaboom.
Its kind of like automobile gas stations... it would be nice to never have a leak and LARP that we'll be very serious and studious about never having a petroleum leak, but that mindset leads to design and operations decisions which become environmental disasters when the inevitable happens. Better to build everything with the assumption the underground tank will leak one liter per month than to plan the leak rate will be zero. Also no level of idiot proofing is perfect; that leak-tolerant station design will still fail if some truck driver pumps 2000 gallons into a 1000 gallon tank while not paying attention.
There's also safety tolerance / clickbait issues. Natgas averages about ten customers dead per year in the USA. Its very difficult to get stats for in-house mis-adventure but it seems about three hundred die per year from in-home electrocution, so its kinda a rounding error. Food related illness (food poisoning leading to dehydration leading to death) kills about 5000 per year in the usa per google (which seems high?) so you're about 30 times more likely to electrocute yourself at home than die in a natgas explosion, and 500 times more likely to be killed by whats in your cooking pot than whats heating your cooking pot. Given that, more investment to prevent natgas deaths would likely result in more deaths from other causes; taking money away from "stop drinking corn syrup" or "cook meat to 165F" will kill more people than a 10% improvement in long term natgas death stats would save.
It does, but those will vent to the atmosphere, if there is enough overpressure and it lasts for long enough the amount of gas released would be considerable.
It seems one solution might be to have a small constantly-burning flame (the other posts here that say small gas leaks are basically being ignored suggest that it wouldn't really cost much more) at the relief valve exits, so that if they vent, the excess gas just gets burnt. Obviously the relief valve exits should be located away from flammables, and preferably visible so that extended overpressure events are immediately noticeable from the flames coming out:
The goal is to not silently create an explosive mixture, but to burn vented gas to make it harmless and create a visible indicator of extended venting.
> It seems one solution might be to have a small constantly-burning flame (the other posts here that say small gas leaks are basically being ignored suggest that it wouldn't really cost much more) at the relief valve exits, so that if they vent, the excess gas just gets burnt.
When the amounts are larger that is exactly how it is done. But for trace amounts the mixture would not reach a concentration level where it is combustible and then you need a whole setup around that to batch it to the point where it is.
Hard to do that passively, also, you could no longer do that indoors.
In legacy “low pressure” systems, regulators aren’t used at the residential end at all. Our neighborhood was recently upgraded to a “high pressure” system, and regulators were introduced, along with vents cut through the foundation to the outdoors.
Our neighborhood was built in 1950. This system upgrade included replacing all the mains in the road as well.
