It turns out that the rolling electrical blackouts in the mountains of Colorado on December 30, which were a consequence of the devastating Marshall wildfire in Boulder Country, Colorado, were about the gas company avoiding having to relight pilot lights!
If there is anything that the gas company hates doing, it is having to send out workers to go from door to door to relight everybody’s pilot lights.
When I was a child, our house had two pilot lights. One was in the furnace, and the other was in the water heater. Every now and then it was necessary to relight one or another of the pilot lights. Somehow three or four times a year, the furnace pilot light would “go out”. Maybe a gust of wind came down the flue from the roof and blew it out. Invariably this problem of the pilot light “going out” happened on some very cold night in the middle of winter. Anyway from age ten I was the man of the house so it was up to me to go down to the basement and lift off a metal cover from the furnace and then unscrew a couple of wing nuts and pull away a metal door and then I could see a pilot light nozzle back in a dark place. There was supposed to be a faint blue flame at that nozzle. But no flame was there.
I would rotate a gas control valve to a particular position. I would then light a stick match and position its flame near the nozzle. I would then push a red button to permit gas to flow to the nozzle, making a very faint hiss, and as often as not, this would extinguish the flame on the match.
On a second or third try I would manage to get the pilot light burning. This would warm up a heat sensor which would then eventually set a latch at the gas control valve. This would permit me to cease pushing the red button that I mentioned earlier. The only way to know whether the latch had or had not gotten set was to release the red button and watch the pilot light to see whether it kept burning or whether it stopped. If I had released the red button too soon, then this would call for yet another try.
Eventually after a couple more tries I would manage to keep the pilot light burning.
The next step was to rotate the gas control valve around to its original position, permitting gas to flow to the main burner, which it would do if the house thermostat happened to be calling for heat at that moment. Which of course it usually was since the whole reason I was down there in the cold basement was that the whole house was cold and we figured the explanation was probably that the pilot light had “gone out”.
So when I would rotate the gas control valve around to its original position, the usual next event was that the main burner would burst into full flame with a bright dramatic whoosh. This was of course very satisfying because it meant that within a few minutes, a temperature sensor switch in the furnace would sense the warmth from this burner and the sensor switch would turn on an air blower, and warm air would get blown upstairs and my mother and sisters would get warmed up. And later I could go back upstairs from the cold basement and rejoin them.
Seeing the main burner in full flame, I would replace the metal door and its two wing nuts, and slide the metal cover back into place, and put the box of stick matches back at its home on the top of the furnace until the next time it would be needed.
What are the chances I would have blown up the house? Pretty much zero. The things that the homeowner can do with the kind of pilot light that has a button and latch are nearly impossible to get wrong in the direction of blowing up the house. There is no combination of mistakes that the homeowner can make that will lead to unburned gas flowing out into the house and presenting an explosion risk. It is, however, quite easy to get it wrong in the direction of failing to get the furnace running again. When this task fell to me during my childhood, I nearly always had to try two or three times before I could get the big burner going again. I would usually go through at least two or three stick matches before I succeeded at getting the furnace back into service. But no, there was basically no risk of my blowing up the house.
Nowadays with any modern gas-fired appliance this business of having to “relight the pilot light” never even arises. This business never arises because there is no pilot light and thus it never has an opportunity to “go out”. In modern-day gas-fired appliances, the way that it works is that there is a “spark ignition”. If a house thermostat calls for heat, the a sophisticated electronic controller follows a predetermined sequence of steps. The controller opens the gas valve to supply gas to the burner, and generates a spark to light the burner. A sensor permits the controller to learn whether or not it succeeded in lighting the burner. If the controller did not succeed, the controller closes the gas valve for some time to allow any gas in the burner to disperse, after which the controller tries again. If after a couple more tries it still does not succeed, the controller actively chooses to “give up” and it will not try any more until somebody resets the controller.
There are some really old gas appliances that have a pilot light, but that do not even have the button and latch that I described. In those really old gas appliances, the flow of gas to the pilot light nozzle is “always on”. If the pilot light happens to be lit, which of course it usually is most of the time, then that is fine. But if the pilot light happens to “go out,” then there will be a very small trickle of unburned gas flowing out of the nozzle into the room. By far the most common outcome from such a trickle of unburned gas is that eventually somebody walks into the room and smells the rotten-egg smell and realizes that the pilot light needs to be relit. And they open a window or door and allow the gas to dissipate and then they relight the pilot light and then they close the door or window and then everything is back to normal. Or in a rare case there happens to be some open flame somewhere in the room, and a long time passes with nobody walking into the room, and the errant gas reaches the flame and makes a little “poof” that startles people but is harmless because the gas had gotten diluted to the point of harmlessness as it spread. Or in an almost vanishingly rare case the trickle from the pilot light actually manages to accumulate enough to be dangerous. In one instance in a million, a house blows up. This sequence of one unlikely thing followed by the next unlikely thing is quite rare, but of course when it happens it is all over the news, and people sort of assume that every pilot light in every gas appliance in every house is likely to be next. There are people who actively say things like “I prefer my all-electric home because I don’t want the risk of a gas appliance in the house”. Such people are perhaps unaware that the risk that they are trying to protect against applies only to types of gas appliances that you cannot purchase any more and indeed could not have purchased even in the past fifty years. Such people should probably also avoid ever stepping off the curb to cross a street since it is impossible to rule out the possibility that one might get run over by a truck. But I digress.
Which brings us back to the situation that if there is anything that the gas company hates doing, it is having to send out workers to go from door to door to relight everybody’s pilot lights.
What is it that happens that puts the gas company in the situation of having to go around and relight pilot lights?
The thing is that normally the gas company is providing the gas to everybody’s homes in a more or less continuous way, at some more or less standard amount of gas pressure. This keeps the gas flowing to the pilot lights. There is no choice, however, but to consider what might be the consequence if the gas were to stop flowing and later if the gas were to be restored to normal pressure. A related event would be a partial loss of pressure in the gas distribution system, followed by a restoration of pressure in the gas distribution system.
For a gas appliance with a spark ignition, by which we mean any gas appliance purchased within the past twenty years or so, neither of these events presents any safety problem. Yes if the gas were to stop flowing completely, the gas appliance would not be able to do its job. It would not be able to heat the water, or heat the house, for example, until such time as the flow of gas was restored. If the pressure were to be reduced and later restored, the appliance might not work right or might not work at all during the event. But nothing about these events would present a risk of unburned gas leaking out into the house. The spark-ignition appliances are designed in just the way that you would hope they would be designed, so that either they successfully burn gas and do whatever they are supposed to do, or they shut off the gas. But there is no in-between behavior in which unburned gas is flowing in an uncontrolled way into the house. Eventually when the event is over, the gas appliance is again able to do its job.
The gas appliances that predated the spark-ignition appliances were the ones with the button and latch for the pilot light. These are the ones where as a ten-year-old boy I would hold down the button and try to light the pilot light and maybe the latch would get set and if it did get set, the pilot light would stay on. As it turns out, these gas appliances are also completely safe even if there is one of these gas outage events. If the gas goes out for a while, or if the gas has a pressure drop for a while, the consequences will not be a flow of unburned gas into the house. The consequences will be that the pilot light goes out, and the latch releases, and no gas can flow at all. And yes this means the homeowner will not have hot water. Or the furnace will not heat the house. Things will be cold until somebody relights the pilot light. But there is no safety problem. No unburned gas leaks out. There is no in-between behavior in which unburned gas is flowing in an uncontrolled way into the house. Eventually when the event is over, the house is cold or there is no hot water. And somebody is going to have to relight that pilot light. But there is no safety risk as a result of the event.
So the safety problem that we are talking about is that some small fraction of very old gas appliances in some very old houses date from so long ago that (a) they do not have spark ignition and (b) they do not have the button-and-latch control on the pilot light. Such an appliance might present a safety risk if the gas goes out and then gets restored to service. Such an appliance might present a safety risk if the gas pressure drops too much then returns to normal. Some very old ovens are like this. Some very old ranges (gas cooktops) are like this.
I’d guess that there are entire neighborhoods where not a single house presents any risk of this kind. Any housing subdivision constructed in the 1950’s (or more recently) will have been built at a time when the building codes required every gas appliance to have at the very least a button-and-latch control on the pilot light. Any housing subdivision constructed in the last two decades will have been built at a time when the only kind of gas appliance that could be installed would be one with spark ignition. But I suppose one cannot rule out the possibility, however remote, that some homeowner even in a very recently constructed house might have gone antique shopping and might have purchased some old stove and range of the kind that has an always-on uncontrolled pilot light that does not have any button-and-latch control. Such a stove-and-cooktop, if placed into service, would have a pilot light that could “go out,” and if it were to “go out,” then unburned gas would flow into the room.
So now we return to the Marshall fire of December 30 in Boulder County, Colorado. This was a big fire and it damaged several important gas transmission lines.
In most of Colorado, there is a single company that is both the gas company and the electric company. It seems to me that this situation should not be permitted to happen because it is surely an antitrust violation. But nobody asked me. It turns out that across the US, since the 1950’s, in most states what has happened is that consolidation has taken place so that in many service areas, the same company is both the gas company and the electric company. I guess the thing that nominally keeps it from counting as an antitrust violation is that there is a state regulator that keeps a close eye on everything to ensure that the consumer is protected completely. And yes I am sure that every state regulator does accomplish that goal at complete arm’s length from the utilities that it regulates. It must be so. But I digress.
Anyway, in most of Colorado there is a single company that is both the gas company and the electric company. The company is called Xcel Energy.
On the evening of December 30, Xcel Energy noticed that it was losing gas pressure in its gas distribution system. This was due to the Marshall fire in Boulder County, down at an elevation of 5000 feet, on the east side of the Continental Divide. The parts of the gas distribution system that were losing pressure as a result of the fire were, however, up in the mountains at elevations of 7000 to 9000 feet, on the west side of the Continental Divide. I don’t know exactly why this would be that the events down at 5000 feet on the east side of the Continental Divide would affect the gas distribution service areas up at 7000 to 9000 feet on the west side of the Continental Divide, but that is what Xcel Energy said in its news releases. I guess the supply comes from down in the flatlands and I guess the distribution network up in the mountains is completely dependent upon the supply systems down in the flatlands.
So from the point of view of Xcel Energy, the concern was that the pressure in its mountain service areas might drop enough to allow those old-fashioned pilot lights to go out. If that were to happen, then later when the pressure came back, those nozzles of the old-fashioned pilot lights would permit unburned gas to flow into homes. This might lead to houses exploding, and then Xcel Energy getting sued.
So I guess Xcel never quite comes out and says this, but it must be they have a rule that if their distribution pressure gets low in any particular service area, they have to shut off the gas completely in that service area. And then they have to restore the gas service in that service area only as part of a door-to-door process of relighting everybody’s pilot lights. They cannot permit the pressure to get low and then return to normal. If it gets low, they have to shut it off completely because there is no telling whether maybe there will be a few of those old-fashioned pilot lights out there that would go out. And Xcel Energy cannot assume that the homeowner is capable of handling this situation safely. A trained technician has to be the person who relights the pilot lights, says Xcel Energy.
Probably they have to visit every home at least twice. First they have to go around to every house and turn off all of the gas meters at every house in a given service area. Then they can turn the gas back on for that service area generally. Then the trained technician goes back to every house, one by one, and knocks on the door to see if somebody is home. If somebody is home, they can get inside and relight the pilot lights, coordinating this with turning the gas meter back on for that house. Now of course about 80% of the time this visit in the house turns out to be a complete waste of time because every gas appliance in the house has spark ignition so there is not any pilot light that needs to be relit. So if only they could have known this, there would be no need to have turned off the gas meter for that house and later turned it back on. They could have simply restored the gas service to that house at the earliest possible moment and not needed to do two truck rolls.
Not only that, but half or two-thirds of the time nobody will happen to be home when the trained technician goes out and knocks on the door to see about relighting the pilot lights. So they leave a door tag and make a check mark on a list on a clipboard. This house has by now had two truck rolls but it will now require a third truck roll.
Yes 80% of the time, there is actually no need for any of this business of two or three or four truck rolls because every gas appliance in the house has spark ignition.
Another 19% of the time, there is still no safety-related need for any truck rolls because even if the appliances were not so modern as to have spark ignition, at least they would have the button-and-latch controls on the pilot lights. The homeowner can be completely trusted not to blow up the house if the pilot lights have the button-and-latch controls. But I suppose in fairness it can be said that not every homeowner will be successful at getting the pilot light relit. So maybe the truck roll will be important just to get the pilot light going again. But if so, this does not require shutting off the gas meter in a first truck roll and then doing a second or third truck roll to relight the pilot light. A single truck roll should suffice. And anyway this is not for safety reasons. It is not to keep from having a house explode. It is simply to help out the homeowner who cannot figure out how to follow the instructions on the big yellow label on the side of the appliance that explain how to relight the pilot light.
Which leaves us with the 1% of houses where there is an uncontrolled pilot light, the kind where if the gas drops and later comes back, it will lead to unburned gas flowing into the house. It is probably not even 1%. Probably only a fraction of a percent. But yes there must be a few. Probably fewer such houses with every passing year, but they cannot be all gone even now in 2022. (Yes, now the year is 2022. Yikes! But I digress.)
So what did the gas company do when it realized that the gas pressure in its distribution system in the mountains of Colorado was dropping below nominal levels?
It realized that if the drop got worse, eventually that 1% of homes in the service areas of low pressure would have their uncontrolled pilot lights go out. At which point, the door-to-door relighting process would have to be carried out for 100% of the homes in the service areas of low pressure.
What the gas company did was, it took advantage of the fact that it is also the electric company. The gas company conducted a series of rolling electrical blackouts in its various mountain service areas. Here is how the local newspaper, the Summit Daily News, described it:
According to a news release from Xcel, the wildfire impacted its system that supports Summit and Grand counties. To decrease the amount of natural gas being used by furnaces and to keep the system up and running, the company’s release said it was going to conduct periodic electric outages for customers in Summit, Grand, Lake, Eagle, Saguache, Rio Grande and Alamosa counties for six to eight hours.
Though the release and an alert from Summit County said these outages were to last about an hour, some community members experienced outages for as short as 30 minutes and as long as eight hours, said Brian Bovaird, Summit County’s director of emergency management.
[…]
“The issue was actually not the power grid; the issue was the natural gas pipelines,” Bovaird said. “So when all of those houses burned down, a lot of gas lines and feeder lines got compromised, and so obviously, there is a lot more natural gas running through those lines, so there’s almost instantly a high demand on the system for natural gas.”
Bovaird said the worst-case scenario was that gas pressure would decrease throughout the Colorado service area to the point where customers lost pressure to their homes, causing pilot lights to go out. This is a major cause for concern because when they go out, gas keeps coming out of a pilot light and could collect in a home, causing risk of an explosion. Pilot lights are used in appliances like furnaces, gas fireplaces and hot water heaters.
“The biggest hazard is if there still is enough gas pressure coming into the house but the pilot light was extinguished, then people’s houses could fill up with gas, so obviously, that’s a super dangerous situation that I think Xcel would say would be their primary goal for this,” Bovaird said.
As to why rolling blackouts were needed, Xcel’s latest release stated that the blackouts were put in place to decrease the amount of natural gas being used by furnaces so the natural gas system could continue to provide service to customers.
So let’s put this in plain language. The gas pressure in the gas distribution systems in Summit and Grand counties (two counties) was getting to be below its nominal level. This meant that the 1% of homes in Summit and Grand counties that had extremely old gas appliances might have their uncontrolled pilot lights go out.
To prevent this from happening, what Xcel came up with was the idea of cutting electrical power to its electrical customers in seven counties, so that the electrical customers would be incapable of making use of their furnaces. These are the “rolling electrical blackouts” in the mountains of Colorado that were all over the news on the evening of December 30. This would then make it so that the furnaces would not be able to draw any gas from the gas distribution systems in those seven counties. This would then permit the gas pressure in Summit and Grand counties to stay close enough to nominal levels to minimize the risk of the uncontrolled pilot lights in those 1% of homes from “going out”.
Talk about the needs of the few outweighing the needs of the many.
There would be two ways to handle the pressure drop in Summit and Grand counties. One way is, let it drop and let it come back. The problem with this is, there are the 1% of homes with uncontrolled pilot lights that might lead to a house exploding and then Xcel Energy getting sued.
So that is out of the question, I guess. Because those 1% of homes exist, Xcel Energy probably has a policy that once the pressure gets low in a given service area, Xcel has to cut it off completely and not turn it back on. Then Xcel has to go out and shut off all the gas meters, door to door. Only then can Xcel turn the gas back on for the service area. Then Xcel has to go out and get into each house, one by one, to turn each gas meter back on, one by one, coordinated with relighting the pilot lights in the house. At least two truck rolls to each house in the affected service area.
Which would cost Xcel Energy lots and lots of money. And, I guess, would inconvenience the gas customers in the affected service area.
To save all of that money for Xcel Energy, electrical customers in seven counties had to endure rolling electrical blackouts.
So if you were to ask what it was that Xcel Energy sort of hoped it might accomplish in terms of a high-level goal, it was that it hoped to convince most or nearly all of its customers in a seven-county area to not turn on their gas furnaces. So what one can imagine is Xcel could just drop everybody an email that said “hey if you have a gas-fired furnace could you please just turn it off for a few hours?” And if the homeowners had all cooperated by turning off their gas-fired furnaces, then this would have achieved the high-level goal.
Of course there is no way that this “asking nicely” way of controlling the gas-fired furnaces would have worked. So instead Xcel Energy used a hammer to swat a fly. It had a very crude but effective remote-control way to turn off all of those gas-fired furnaces, namely by cutting off all of the electrical power to the houses that contained those gas-fired furnaces.
This remote-control approach for turning off furnaces also of course turned off the furnaces in the homes in this seven-county area where the furnace was fueled by a propane tank rather than being fueled by gas from Xcel. And of course some of the homes in this seven-county area make use of electrical baseboard heating. This remote-control approach also turned off the electrical baseboard heat in all of those homes.
I suppose some of the homeowners in the seven-county area that were affected by the rolling blackouts were reliant upon electricity not only for a furnace but also for other things such as lighting or the ability to open and close their garage doors, or keeping food from spoiling in their refrigerators.
What should be done about all of this? Maybe Xcel Energy should go door to door to locate the handful of gas appliances that are still in service that have uncontrolled pilot lights. There cannot be even a thousand homes in Xcel’s entire customer base in Colorado that have such ancient appliances. What is the cost of one set of truck rolls for the relighting of pilot lights in a typical service area? I’d guess it to be something like a quarter of a million dollars. Xcel could offer to replace every such appliance and I bet it would not even cost a quarter of a million dollars. For every home in which such a replacement got accomplished, Xcel would reduce greatly the risk of getting sued. The local first-responders would be at less risk if they ever had to respond to an emergency call to the house. The homeowner could feel good about having a newer and safer and probably more energy-efficient appliance. This would be the kind of program where nobody ends up worse off, and everybody wins.
Interestingly enough, the clip boards carried by the folks who went door to door for this most recent problem most likely collected on their clipboards data that identifies each home with an ancient pilot light.
It would perhaps be cheaper for the utility to extract and save that data in some convenient place and then bribe homeowners to upgrade their gas appliances.
Perhaps the public utilities commission might be convinced to pass a rule that allows the utility to turn off and lock gas cutoffs connected to homes with offending gas appliances.
Wow that would be smart. I guess you are assuming the gas company would be smart. Maybe so.
So then in a strange way the gas company should be glad about the opportunity to do the door-to-door because it provides the opportunity to gather this information about who has which appliances that are how ancient.
Some years ago, my natural gas utility shut off the gas supplied to my home, as an incident to replacing an underground gas line due to a leak somewhere, and they restored the gas service without the practice you described where they would require a technician to relight the pilots.
The utility left a door tag inviting me to call to arrange to have the pilots relit, but I did it myself.
Perhaps best practices have changed over the years so as to require the technician visit as a condition of restoring service. Or perhaps codes in my location forbid the ancient, unmonitored pilots. Alternatively, it wouldn’t surprise me if some law here relieved the utility of liability because IMO, utilities here own the legislature and the PUC and they get whatever they want.
Also, in my experience, the thermocouple-based pilot valves are extremely reliable and long-lasting. I think I’ve had to replace one thermocouple in several decades of use. Especially compared to software-controlled and cost-reduced devices of modern vintage, which seem to be incredibly brittle, it’s amazing how well the old pilot valves work. I salute whoever designed and built them.
I also remember the hassle of re-lighting pilot lights. In one case, the person who built the house had installed the water heater in a 12″ deep well in the floor of the basement, with a drain. Clever idea if the heater leaked. But it made it really difficult to reach the pilot light. Extra long matches help, but I learned to use one of those BBQ grill lighters, which have a nice long reach, an adjustable flame, and don’t lead to burned fingers.
I used the same trick in a house I rented from 2006-2012 where high winds would regularly blow out the pilot light. That area is known for high winds. Until a few days ago, it was in the Sagamore subdivision in Superior. Now it is gone.
Also,
1. I wonder what Xcel and its brethren will do once a substantial number of customers decide that utility-supplied electric power is not sufficiently reliable–perhaps in part as a result of this very incident–and therefore install on-premise storage or generation, thereby depriving Xcel of remote control of gas demand via rolling electricity blackouts.
2. Regarding the suggestion that appliances newer than twenty years or so have spark ignition, the threshold would appear to be a few years later. I have a gas appliance installed only 15 years ago that has a conventional thermocouple-countrolled pilot.
3. Spark ignition is not the only form of pilotless ignition in gas appliances. I have an appliance with a silicon carbide ignitor (generically, a “hot surface ignitor”) that is made hot (around 400 F) by passing electric current through it; a little stream of gas is exposed to the hot surface. It is much simpler than the spark ignition systems.
As for your point number 1. I live in Summit County, which is one of the counties where Xcel imposed the rolling blackouts. In Summit County, I’d guess that at least ten percent of the homes have backup generators to keep at least the essential house systems running during a power failure. Perhaps ironically, many homeowners in the county have their backup generators set up so that the generator is powered by … wait for it … the Xcel natural gas service to the house.
This is likely also the case in the other six mountain counties where Xcel imposed rolling blackouts.
My house does not have a backup generator. Instead, my house has 27 kilowatt-hours of battery backup that switches into service automatically in the event of a loss of power from Xcel energy. The battery backup feeds ten kilowatts of AC inverters so that up to ten kilowatts of loads in the house can be powered by the battery backup. In addition, my house has sixteen kilowatts of solar panel inverters. These rolling blackouts happened at night, so the solar panels were not of any help during this particular time.
The point being that Xcel was not able to remotely control the natural gas consumption in my house. Which reinforces your point number 1.