Explosives

Good to hear that there is nothing wrong with the hardware.

Comparing this incident with a stage falling over and killing and injuring people is quite unfair. The safety measures worked, no one got injured or killed. The safety are there to protect the audience not the show.
This is a human error and it is bad for business, but all the necessary safety measures were in place. It was a bit clumsy and it makes the company look bad, but it was not unsafe.(Based on what I have been able to find out)

This can not in the future be avoided by another standard. Perhaps for this company can have a new internal way of working that will avoid this happening again.

Asking how long a delay fuse you can have is like asking how long an extension cord you can have. Cut it shorter or longer. How long they normalay are depends on tradition. If you use a lot of Italian product such as roman candles it would make sense that they are at the longest point approxiamtly as long(in time) as the candle. Usually between 20 to 30s.
 
actually, a standard of checking your control program (which in this case fired both the backup and the main file) would be needing to be in place. they checked the main file, and they checked the backup, but failed to check their launch program. it can, so when you tell us no it can't. I claim bull honky.

As for safety measures, the show didn't stop as soon as they knew something was wrong. Hence safety wasn't enacted. They also fired 5 mins early. So it may not have been all clear. Safety was breached although it was fortunate no one got hurt. This could have ended very differently.
 
but failed to check their launch program.

From what little I know and have read, this seems to be the case.

As for safety measures, the show didn't stop as soon as they knew something was wrong. Hence safety wasn't enacted. They also fired 5 mins early. So it may not have been all clear. Safety was breached although it was fortunate no one got hurt. This could have ended very differently.

The show couldn't stop at this point. The cues fired at the start of pre-roll instead of the usual test sequence happening. I'm new to this end of the business, but what I do know is that everyone is to be all clear once pre-roll is started (for this very reason). The press release states this, as it is also common practice. Once the system is armed [which has to happen before pre-roll] everyone is to be all-clear.
 
Good to hear that there is nothing wrong with the hardware.

Comparing this incident with a stage falling over and killing and injuring people is quite unfair. The safety measures worked, no one got injured or killed. The safety are there to protect the audience not the show.
This is a human error and it is bad for business, but all the necessary safety measures were in place. It was a bit clumsy and it makes the company look bad, but it was not unsafe.(Based on what I have been able to find out)

I am quite thankful that all the safety measures in place kept everyone from harm. I think you may have misunderstood some of the comparisons I was attempting to make between the two large scale failures. One, human error is often involved. This is why pilots are forced to manually go over a pre-flight check list even if they have thousands of hours of flight time. Humans make mistakes. Two, with every large scale failure, after the investigation, we learn how to prevent this type of failure in the future. Sometimes we have to write new standards. The safety measures put in place for outdoor stages are also to protect the audience/crew, not the show. If a show is a flop because of bad weather or something else out of control, then so be it; just as long as everyone goes home in one piece.

This can not in the future be avoided by another standard. Perhaps for this company can have a new internal way of working that will avoid this happening again.

Asking how long a delay fuse you can have is like asking how long an extension cord you can have. Cut it shorter or longer. How long they normalay are depends on tradition. If you use a lot of Italian product such as roman candles it would make sense that they are at the longest point approxiamtly as long(in time) as the candle. Usually between 20 to 30s.

This is where I don't understand. When you guys are talking about fuses, I don't know the difference between what you are using and that on common fireworks sold to the public. When the system is automated, is it a different type of ignition system? When there are multiple fireworks that are set off with one cue, is it just different length fuse or is there some sort of timer that is triggered in the computer sequence that says fire off A then B then C when cue one is started?

A new standard could be written if the old one is broken. However, if it can safely fail as this did, there may not be cause to find a different way to do it. I would think that the industry would want to potentially save a part of their show if they could somehow stop a sequence from continuing when it was accidentally triggered. Really, I am not being a troll here, just trying to understand.
 
Wow. Seriously guys, this is making my head hurt. You don't know enough to even have this discussion, but are making some wild claims.

With that said, I will continue to try and explain things and answer questions. At the same time, you really do need to take 10 steps back and understand about how the fireworks function prior to making claims on how to improve them. As San Tai said, there is NO safety violation or practice that would have prevented this.

1) The technical part of the press release doesn't make sense to those who know the firing system they were using. That's just not how it works. Files aren't merged together and the software didn't screw it up. The firing system manufacturer said he didn't even understand what they were actually saying. They edited their own file on site and didn't look at it before loading into the panel. No backup file like this would have messed it up, either. The error really was as simple and dumb as it gets. Think of a sound guy playing the wrong mp3 for a big event. All he had to do was listen to it before playing. Should the government require that all sound people now double check before hitting "play"? I realize it's not a good analogy, but work with me here...

2) Fuses. Have you ever bought a string of firecrackers? There is one ignition and then a fuse runs along the middle. Each firecracker individually has a fuse that touches the main run. Thus, you light the string and can't stop it until it burns out. The finale shells in a show are often stringed together (called finale chains). The end of the chain is lit by an electric match (ematch or igniter), and the entire chain goes. Sometimes these have delay fuses to maximize the time of the chain and might last up to 10 seconds or so.

They also had what are called "cakes." These are like the preloaded boxes you could buy at a fireworks stand. Once again, you light the fuse and it shoots a sequence of tubes internally. Once started, you cannot stop it. These typically last 5 to 30 seconds, but I've seen them last for up to 60 seconds. As San Tai said, a fuse can be as long as you want it to be...


Since I don't know about the use of timed chains with delay fuses or standard quickmatch chains, how long of a sequence is standard for a single firing sequence? What I am trying to understand is with this accident which took over 30 seconds to fire all of the fireworks and all the sequences were triggered to start at once, how long of a chain can be set for a single trigger? How do these sequences get their commands in a system like this? Does it send one command to a single electronic match which in turn triggers additional matches (or equivalent) or are they set up like an effect sequence where you have a start command for the effect, but then there are individual steps in that effect which are individual commands?

You're overthinking it. Read above on chains. One end is lit, and the the fireworks go. Although this doesn't sound safe because you don't have control, it's been this way forever. It's just how fireworks are made. Don't tell us that it needs to change because it won't. That would be like the government outlawing par cans. They're THAT common.

I disagree that they only needed to have the last paragraph in their explanation of how events occurred. I think that it was very upstanding of them to disclose everything that they did. I don't think that we can come to begin to understand everything that goes in to putting on a show like this. By only putting in "it was human error" is like saying that everything that occurred last summer in Indiana was that there was a wind storm. We belittle ourselves to say that we can boil down a situation to something so simplistic and thus perpetuate to the public that what we do is simple as well. Sure, it was an avoidable error. I believe that this company will probably help set the standard to make sure that this error is not repeated, especially on this caliber.

I was agreeing with everything until the last sentence. It's like you guys want there to be more to the story than there is. Again, they made a very simple (and dumb) mistake. I was sure it wouldn't have been this simple, but I was wrong. Point is, there isn't going to be some new "standard" or ensuring this won't ever happen again. People will always accidentally push a wrong button or make a dumb mistake.

actually, a standard of checking your control program (which in this case fired both the backup and the main file) would be needing to be in place. they checked the main file, and they checked the backup, but failed to check their launch program. it can, so when you tell us no it can't. I claim bull honky.

As for safety measures, the show didn't stop as soon as they knew something was wrong. Hence safety wasn't enacted. They also fired 5 mins early. So it may not have been all clear. Safety was breached although it was fortunate no one got hurt. This could have ended very differently.

I agree, everyone should check their program before loading it. It goes without saying. There won't be a government standard enacted for that. Internally, yes, they need to figure something out. Not sure what you mean by "it can, so when you tell us no it can't. I claim bull honky." You can check the program. I don't think anyone said you can't?

This is where I don't understand. When you guys are talking about fuses, I don't know the difference between what you are using and that on common fireworks sold to the public. When the system is automated, is it a different type of ignition system? When there are multiple fireworks that are set off with one cue, is it just different length fuse or is there some sort of timer that is triggered in the computer sequence that says fire off A then B then C when cue one is started?

It's different than what's sold to the public but same principle. On professional shells, the "fuse" is called quickmatch which burns approximately 70 feet/second. There are also fuses that can be inserted between pieces of quickmatch to make each shell last a little longer. These are called time delay chains. Although a chain of shells is lit, the entire sequence to shoot all shells may last 5-10 seconds. Very popular in Europe. When fireworks are lit electronically, it doesn't matter what type of system is being used. The underlying idea is the same to pop the electric match. At "0" every single e-match in the show fired at once. All effects then shot. Some lasted longer than others. It's that simple.

A new standard could be written if the old one is broken. However, if it can safely fail as this did, there may not be cause to find a different way to do it. I would think that the industry would want to potentially save a part of their show if they could somehow stop a sequence from continuing when it was accidentally triggered. Really, I am not being a troll here, just trying to understand.

Yes, we all know a new standard COULD be written but nothing is broken. What would it say? Look at your firing script prior to loading into the panel. Everyone does this, someone forgot to. A new standard won't fix someone taking a shortcut and messing up their show.

Please ask questions, then make suggestions. It's hard to even discuss because you're making demands for more safety and better standards before you know how things work. Hopefully the above makes it easier to understand what happened. If not, I'm happy to keep answering questions :)
 
Since procedural mistakes happen, could the firmware in the firing system perform an integrity check on the timeline it is given, forcing the user to review flagged areas with unusual cue density, duplicated cues, etc?

Or would that produce too many false positives that operators would just skip through it...
 
Since procedural mistakes happen, could the firmware in the firing system perform an integrity check on the timeline it is given, forcing the user to review flagged areas with unusual cue density, duplicated cues, etc?

Or would that produce too many false positives that operators would just skip through it...

Good thoughts. It's been discussed and would be completely up to the firing system manufacturer. It's one of those cases that most don't see the need for a "fix." It was an honest mistake on the user. There's only so much you can do to help that. Sometimes you do intend to shoot all cues at once for particular effects. It would be hard for the computer to know the human intention.

It should be noted that this firing system is arguably the most widely used in the word. Shooting several shows a night, year round. This was a small, but dumb, mistake but happened to be on a big show. Just because it was on a big show doesn't mean things have to change. In fact, the firing system performed as it should (and pretty well being able to shoot so many cues at once). I'm sure this has happened before, but nobody knows because it wasn't a large public show. If this was happening regularly, the industry would cry out for a fix.
 
okay, now that we are understanding it a bit more, What about the electronic fuses? I've seen quite a bit of them. Sure they are a bit more expensive, but seem to have much more control. I was more saying a standard could be written for the above. I wasn't expecting a government mandate standard, but rather an industry wide. Similar to my area when anyone is working overhead, its not mandated that we lay down carpet under the work area (on the grid) but we do it because it helps prevent pins, shackels, and other hardware and tools from falling down below the grate.
 
I'm not sure what you mean by electronic fuse? An ematch (electric match or igniter) creates a small spark when it receives electricity from the firing system. They were using these...It's the only option when firing electronically and they are the standard...
 
couldn't you place those directly in the charge, or shell? I know there is at least one manufacturer placing an wirelessly controlled e match in their shells to time the explosion right at the set time instead of relying on a "analog" fuse.
 
They were placed on the shells. For some items one ematch is used (chains and cakes). I'll use the firecracker example again. If I wanted to light firecrackers electronically (and I have), I use one ematch for a roll of 16k. You're suggesting the equivillant of placing an ematch on EVERY single 16,000 firecrackers. Not gonna happen :)

A finale chain consists of multiple shells fused together. You light the single fuse and all others are fired at that point. For example, you might only fire off 80 ematches in a finale, but light 900 shells at once. Those chains might have time delay fuses which will take each chain 2-10 seconds to finish.

Couple things about the timed shells...

You wouldn't control each individual shell in a finale, only the main body.

Yes, there is a manufacturer who had been working on a remote burst charge. That's another technology in itself. No standard firing system can control that. BUT, think of what that means...shell goes up, shell must receive a wireless command to burst. What if it doesn't get the command, what if it is delayed (and comes back down, then bursts). Way more issues with that.

There's nothing wrong with the pyrotechnic fuse that connects chains. That's just how finales and other items are fired and always have been.
 
I really didn't want to post this video, but this probably shows it better than anything. This is not me, my company, or my video. Safety is a little lax in my opinion. They were doing a demonstration for employees. Not using this as a safety video, so don't address that. Just showing you what chains look like.

Here's what you'll see:

Guy manually lights string of quickmatch (fuse) and it hits the chain (see how fast that quicmatch goes!). The firing of the chain takes a lot longer than normal. Typically, as soon as the chain lights, the shells start to launch.

Long Quickmatch to Fireworks Finale Chain - YouTube
 
Again, I did not make myself quite clear enough. When I was mentioning that Garden State would help set the standard, that was not to be meant that they would help write the government standard, but set precedent in the industry. Similarly we have precedents which are "standard" in practice as a good thing to do to make sure things go right.

Based on what you are saying in how the controller works, no amount of additional ematches would have changed a thing except possibly making the whole thing fire even faster. Since the ematch is triggered by timecode and all timecode was set to fire the same time, then having more ematches in replacement of chained fuses would have resulted in a shorter show. The only way to prevent that, with the addition of ematches and shorter chains, would be to have the time code trigger some other series of commands (call it a sub-timecode) where that would take the place of the series of fuses.

I'm not trying to change your industry. That is up to you and others with your expertise if you see the need. I don't believe that just because something has always been done some way that it is necessarily the best way that it can be done. I also don't think the remotely detonated burst is a good idea either (I would be sweating bullets using that one).

I'm glad that we can discuss this now and not think that there are ill intentions. The sad part of the written word is that it is difficult to understand the true meaning without body language and verbal tone. I might try to use more emoticons when I am being sarcastic and it isn't evident.
 
Again, I did not make myself quite clear enough. When I was mentioning that Garden State would help set the standard, that was not to be meant that they would help write the government standard, but set precedent in the industry. Similarly we have precedents which are "standard" in practice as a good thing to do to make sure things go right.

No worries. Garden State followed all industry practices and procedures. They just made a dumb errror in the process. Difficult to prevent the type of mistake they made. It's more of a company by company procedure. Think of it as more of a typo. Sometimes people will make mistakes.

Based on what you are saying in how the controller works, no amount of additional ematches would have changed a thing except possibly making the whole thing fire even faster. Since the ematch is triggered by timecode and all timecode was set to fire the same time, then having more ematches in replacement of chained fuses would have resulted in a shorter show.

Yes. Almost correct. Timecode is just a time base. It's used to sync multiple computers to a precise frame in time. The computer then was told at "0" to fire everything. The computer sends firing commands and sends the electricity to pop the ematch. The timecode just keeps it all in sync.

The only way to prevent that, with the addition of ematches and shorter chains, would be to have the time code trigger some other series of commands (call it a sub-timecode) where that would take the place of the series of fuses.

You lost me...:stumped:

I'm not trying to change your industry. That is up to you and others with your expertise if you see the need. I don't believe that just because something has always been done some way that it is necessarily the best way that it can be done. I also don't think the remotely detonated burst is a good idea either (I would be sweating bullets using that one).

Oh I'm 100% with you on not doing things because they've always been done that way. In fact, that's a main principle that I founded my company on. I've witnessed too many dumb things. However, some things are likely to never change such as how effects are constructed.

I'm glad that we can discuss this now and not think that there are ill intentions. The sad part of the written word is that it is difficult to understand the true meaning without body language and verbal tone. I might try to use more emoticons when I am being sarcastic and it isn't evident.

I'm with you and I can see you're catching on. I'm sure you have good ideas, but you needed to know how it works before making suggestions on what needs to be fixed.
 
Sorry I lost you. What I meant is that a possible solution to a potential uncontrolled long chain (more than say a 10 second chain) would be a different way of controlling a long sequence. Let me put it in a lighting term. We have a cue stack for an entire show. Granted, this is usually a manual cue to cue, but sometimes there is a timecode which drives them. However, we also have things like effects (think chase sequence) which is essentially often built like another cue stack within the main cue stack, which is independent of the main cue stack. An effect could potentially be stopped even when the rest of the cues continue to progress (in case of some need like a bad moving light). I guess I don't totally understand how you are using timecode to build the sequences/fire effects. To me it sounds as if it is being used only to trigger a cue stack (i.e. when you get to X time, fire A, B, and C matches which are fused to fireworks 1-25).
 
Still not getting the comparison. It's just so difficult to compare burning fuse and lights that use cables.

Timecode is just audio beeps and clicks that a computer can read to get a time from. It's totally independent from the firing commands.

A firing system has script installed that says fire X at Y time. If the firing system is being fed timecode via audio cable it will listen for timecode and use that for it's time base. The alternative is just to hit "fire" and start the show using an internal timer.

If it was a single barge out in the water, they wouldn't need timecode. They would just hit "fire" and start the show. Since multiple barges were used, they had one timecode track that was broadcasted via radio and each firing system picked up on it. That's how you can ensure multiple systems are firing simultaneously to 1/100th of a second. It would be impossible to have 5 operators all hit "fire" at the same time.

What is it that you specifically think should be changed?
 
OK, so you are using timecode the same way a lighting console would. So, ignore that. I understand that there's a difference between a fuse and a light with a cable, but what I am driving at is how the command to light/turn on is initiated.

What I would like to understand then is the firing sequence. Let's go by a single barge with a single operator. If this show was a single barge, would there still have just been one fire command for the entire show or would an operator send a command at the start of each song or how does that work? Say I am lighting a dance piece (multiple lighting cues based on music). Since the music is a set time and I know that the lights should always look the same, night to night, I could easily link all of my cues together so that the lights will all change throughout the number. On some beats of music I might have a bunch of lights flash, on another could be a rapid chase, and at the end of the piece is a complex movement of moving light cues. This is done by building cues and effects (triggered by a cue) which are sequenced by a follow (waits and delays) and different things happen due to different commands.

So, if I have this same set-up with fireworks, as I understand it, you have a computer which is preset with an individual timeline (not timecode) for a song where at a certain time from initiation, a match is lit which the fuse may go to one firework (mortar, cake, or whatever) or that fuse may be a sequence of fireworks (one match for multiple charges). The second one is essentially what I am calling an effect on the lighting console except that you cannot stop that effect once it has begun.

What I don't understand is if you can have a sub-set of cues instead of using a long fused effect? Up until this failure, there definitely hasn't been a need, and there may not be a need, I'm not going there. Maybe you could direct me to an online manual for this type of firing system and I could get my answer that way, or at least better phrase it for you to answer my question.
 
OK, so you are using timecode the same way a lighting console would. So, ignore that. I understand that there's a difference between a fuse and a light with a cable, but what I am driving at is how the command to light/turn on is initiated.

What I would like to understand then is the firing sequence. Let's go by a single barge with a single operator. If this show was a single barge, would there still have just been one fire command for the entire show or would an operator send a command at the start of each song or how does that work? Say I am lighting a dance piece (multiple lighting cues based on music). Since the music is a set time and I know that the lights should always look the same, night to night, I could easily link all of my cues together so that the lights will all change throughout the number. On some beats of music I might have a bunch of lights flash, on another could be a rapid chase, and at the end of the piece is a complex movement of moving light cues. This is done by building cues and effects (triggered by a cue) which are sequenced by a follow (waits and delays) and different things happen due to different commands.

No waits or delays. The computer doesn't care what it shoots. Through complex scripting programs, you design a show and it exports a file that is then imported into the firing console/firing system. So the script might look like:

FireTime....Module....Pin
02:15.231...1..........23
02:15.231...2..........23
02:17.108...1..........28
02:19.000...9..........29
03:15.125...4..........03
Etc. for hundreds or thousands of cues potentially

Module is a grouping of individual terminals. Multiple ematches can be fired on a single pin/output. The operator is just monitoring the sequence that is preprogrammed on a computer-fired show.

I'm told the issue was that all of those fire times were at 00:00.000

So, if I have this same set-up with fireworks, as I understand it, you have a computer which is preset with an individual timeline (not timecode) for a song where at a certain time from initiation, a match is lit which the fuse may go to one firework (mortar, cake, or whatever) or that fuse may be a sequence of fireworks (one match for multiple charges). The second one is essentially what I am calling an effect on the lighting console except that you cannot stop that effect once it has begun.

I think we're on the same page. A match lights a pyrotechnic fuse. That fuse burns instantly on a single shell or longer to light multiple effects.

Details but while we're talking terminology, mortar= the HDPE (not PVC!) tube and the shell is the actual firework that is lit. The shell launches from the mortar.

What I don't understand is if you can have a sub-set of cues instead of using a long fused effect? Up until this failure, there definitely hasn't been a need, and there may not be a need, I'm not going there. Maybe you could direct me to an online manual for this type of firing system and I could get my answer that way, or at least better phrase it for you to answer my question.

I have no idea how many cues, but our average July 4th shows had around 576 cues in the 20 minute show. Most had a different time base, some were meant to fire simultaneously. I would have had the same issue if all of my 576 cues were programmed to shoot at 00:00.000. While the manual isn't posted online, you can read more at Welcome to FireOne™ - The Photoshop(r) of the Fireworks Industry!

Chains, cakes, candles, and a slew of other effects have internal fuses that cannot be stopped once fired by the one and only fuse on them. That's just how they're made and come that way from the Chinese (or other) manufacturers. It's just not seen as a safety issue in the industry.

Putting ematches on EVERY single effect or shell would be like saying, you now need to run straight runs of DMX cable to each lighting fixture instead of being able to daisy chain them. Consoles would need hundreds of DMX outputs and all lighting installs would need a LOT more cable. Talk about a nightmare (financially and logistically).

I will say that none of these effects are used indoors or in theater applications. It's just outdoor aerial displays. Much different set of rules for indoor.
 
Last edited:
I will say that none of these effects are used indoors or in theater applications. It's just outdoor aerial displays. Much different set of rules for indoor.
I understand this which is why I am trying to grasp at what you do.
I also understand about the effects with internal fuses. That's kind of like the consumer grade fountains, one fuse, many effects.

I'm sure that I'll come back with more questions after I have time to digest the website you posted.
 
:) Ask away. I do indoor and outdoor shows. I'm happy to help and answer questions. You'll be an expert by the end of this thread!
 

Users who are viewing this thread

Back