5 ms·

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

by tlb 8y ago

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.

jacquesm 8y ago
> 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.
- pas 8y ago
  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?
  - jacquesm 8y ago
    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.
    - sbradford26 8y ago
      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?
      - VLM 8y ago
        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.
- userbinator 8y ago
  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: https://en.wikipedia.org/wiki/Gas_flare https://en.wikipedia.org/wiki/Gas_flare 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.
  - jacquesm 8y ago
    > 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.

by tlb 8y ago

jacquesm 8y ago

> 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.

pas 8y ago
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?
- jacquesm 8y ago
  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.
  - sbradford26 8y ago
    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?
    - VLM 8y ago
      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.
userbinator 8y ago
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: https://en.wikipedia.org/wiki/Gas_flare https://en.wikipedia.org/wiki/Gas_flare 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.
- jacquesm 8y ago
  > 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.