136 - Fire Fundamentals pt 6 - The fire automation in a building

Hello everybody, welcome to the Fire Science Show. I decided that we need to catch up a bit with the Fire Fundamental series and I've decided to record an episode that I wanted to record for such a long time. Today we will talk about what really happens in the building and its fire safety related systems when a fire is detected. I mean, one can think that the role of detection is to detect fire and then everything just should start on automatically, and to some extent that is true. But it's not an accident. It's an effort of work of multiple people, multiple stakeholders and a lot of challenges to get through. A lot of things that must happen and click together. A lot of dangers waiting out there, a lot of issues with potential power drops or power supply In some cases, perhaps too much reliance on the human component, the user, the operator of the system. Really a lot of stuff happening in between the fire being detected and everything in the building standing on its feet, working and actively protecting the people and environment from the consequences of the fire. And a world that I really, really enjoy, because I was given a chance to witness this very closely in hundreds of buildings. Today I'm going to share with you all those interesting things that happen between detection and full operation of the building that perhaps a lot of you would not think about when you look at the fire safety systems in your building. And as the fire science is the science of complexity, I believe there is great value for everyone to understand how this works and perhaps it will change the way how you view buildings or fire safety automation. And after this episode I kind of hope you will view the role of the building user or operator a little differently and perhaps challenge the concept that more complicated systems are better. I'm all into simple solutions because there's the least to break into them. And make sure to wait till the end of the episode, because in the end I'm going to give you, from my experience, the number one issue with all those fire automation systems in the buildings, one issue in their cooperation that literally kills any building in terms of the fire system performance that we've observed in so so many buildings. That simply comes from misunderstanding each other in the process. So if you want to learn that, stay till the end, and for now let's spin the intro and jump into the episode. Welcome to the fire science show. My name is Wojciech Wigziński and I will be your host. This podcast is brought to you in collaboration with OFR Consultants. Ofr is the UK's leading fire risk consultancy. Its globally established team has developed a reputation for preeminent fire engineering expertise, with colleagues working across the world to help protect people, property and environment. Established in the UK in 2016 as a startup business of two highly experienced fire engineering consultants, the business has grown phenomenally in just seven years, with offices across the country in seven locations, from Edinburgh to Bath, and now employing more than a hundred professionals. Colleagues are on a mission to continually explore the challenges that fire creates for clients and society, applying the best research experience and diligence for effective, tailored fire safety solutions. In 2024, ofr will grow its team once more and is always keen to hear from industry professionals. Who would like to collaborate on fire safety futures this year, get in touch at OFRConsultantscom. Okay so, building fundamentals let's go. An important addition to the fire fundamental series, something that when you enter the profession, in some universities, you will be taught a lot, in some universities, you will be taught a little less of, and when you enter the industry, when you enter the practical work of fire safety engineer, you'll learn that you have to relearn this from scratch. So to give it a little bit of structure, in this episode I'll go through detection. We've never talked fire detection and fire alarm systems in the fire science show yet, so I'll briefly introduce you to how we detect fires. Then we will get what this alarm does, how it's received and what's processing it and what happens in that processing, because this is critical in many fires. In many fires where there were fatalities, interruptions in this process were one of the root causes of the fatalities, and we've discussed this with David Burser in the previous episode of fire fundamentals. Then we will talk about how the alarm is transferred to all other devices in the building and actually how other devices are triggered. It's not the magical code that pulls the fire shutter to open. There's a whole automation. There's a whole network between your fire alarm control panels and devices throughout your building that need to execute the commands as they come. And on the commands there's a ton of programming in fire safety which you may have not known or realized how much there is. So I'm gonna touch a bit on fire programming. We will enter ventilation compartmentation, so the execution phases of those systems how do they have to start on and how you can safely turn them off. It's not that simple. I've seen flying doors ripped out of their frames by ventilation system that started abruptly, not in a way like it should. So a very important manner, for sure. And in the end we're gonna talk about supply systems like power supply, water supply and everything that needs to be there to make sure that the systems work. So a lot of things to cover. But first let me tell you why I want to talk about it with you. So perhaps you're already bored with me telling you all the rules in my life. They are ahead in my fire safety career. But here's another one. Since 2013 in the Building Research Institute, my title is the head of fire automation, law and smoke control, and that's a part of fire research department which I lead Not a huge part, just six or seven people, depends on the time but our job is to deal with fire automation. We actually have an accredited electromagnetic compliance, emc and environmental testing laboratory, but that's not the thing that I will discuss in this episode. That definitely is a deep jump for anyone. Here I'm touching on our experience that we've gained by another way of dealing with buildings perhaps less today for me, at least not for my colleagues, and that is the commissioning of buildings with the hot smoke testing method. Almost exactly two years ago, I had an episode on hot smoke testing with my friends from Poland, piotr Smarcz and Janusz Paliszek, which I recommend I'll link in the show notes. Hot smoke tests are commissioning tests during which we really test out everything I am telling you about in this episode. Literally everything that is said in here is put into the test during the commissioning stage of the building, because we must be sure that it operates like the designer expected it to operate. And if we allow a building in which this is unchecked, in which there are errors, in which the devices do not turn on or off in the exact point in time when they should, then the whole fire safety strategy might be violated and the building is not safe anymore. So imagine years of work of the designers, years of work of the construction engineers, years of installing all the systems and, in the end, one error in a programmer's matrix and it's all gone. It won't work. So I feel this is an immensely important role in the building commissioning process to verify all of this, and a very huge responsibility for all those involved to make sure that the things that we put in our buildings, work together in one theater rather than working on their own and not caring much about what's happening around them. So yeah, let's start with the detection. So first, how do we detect the fires? So there are multiple ways you could detect fire in your buildings. The most popular would be through heat or smoke detection. Heat detection means devices that are used throughout the building to measure the air temperature. Pretty much this can be point devices. They look just like smoke sensors. It could be line detectors, linear devices made from glass fiber, or perhaps an array of thermocouples that are spread in the line underneath the ceiling or any other place where you would like to measure temperature, and they continuously measure what's the air temperature around them and once the temperature exceeds a specified threshold, they indicate an alarm. The second way they indicate alarm is when the growth of the temperature is quicker than a set threshold on the first derivative of the temperature over time. So they know how much the temperature has grown and how quickly it did, and through those means you're capable of defining that there's a fire in your building or not. A very common solution, especially in transportation infrastructure. So heat detectors would be how I would detect my fires in tunnels. I don't want smoke detectors in there because they will die because of the tunnel environment very, very soon into their career. So heat detectors are reliable, robust solutions in there, in buildings perhaps used little less, sometimes used as a secondary sensor to reduce the false alarms, but we'll touch on that just in a second. And now smoke detectors, the obvious ones. So in the past we mostly relied on ionization sensors where you would have an ionization chamber. You would measure the ionization current in the sensor and the smoke would alter that. So you would find out if there's smoke or not. Today we are working mostly with optical sensors. Those smoke sensors usually operate in the principle of scattering. They have a small trap inside of them. If you take a sensor into pieces, you will see a trap inside of them, a place where the smoke goes into there's a diode that emits light and there's a bunch of Elements that block the pathway of that light through the sensor. And if the air is clean, the, the light cannot pass through that labyrinth. But where there is smoke inside, the light starts to bump off the smoke particles and Eventually finds out its way to a photo diode that's on the other end of the labyrinth of the trap, and by this we know that there there must be some kind of scattering aerosol in there. Now, is it smoke or not? That's a little more difficult task because you can have all kinds of aerosols that are not smoke in your buildings. So a more advanced detectors would use multiple light wavelengths to test the presence of smoke, because Every wavelength will be scattered in a slightly different manner and by understanding the pattern you can tell if that's a smoke or not, because of how characteristic the smoke particle sizes are in building fires. Another way to Improve their robustness is true, introducing another sensor, perhaps a temperature sensor, perhaps a carbon monoxide sensor in them, a lot of combinations that are present on the market. And, yeah, optical Smoke detectors are pretty much the workhorse of the industry. That's the most common type of a detector. There's also linear smoke detector. Linear smoke detector is used in large spaces and Basically it's a density meter, smoke density meter. You have emitting point, like a laser diode or something, and the receiver that's 20, 30, 50 meters away and there's basically line of light between the emitter and the receiver. When the smoke comes in between them it obscures the light to some extent. So the drop of this transmitters is known to be a signal that the smoke is in between them. There are also other Non-detector based means that help you detect the fire in the building. You can use good old manual fire lamp point. You know the button in the building, their big red button that you should not press normally but in the case of fire you should, and that is pretty much an information to the alarm panel that Fire has been detected manually in the building. You can also, if your building is equipped with sprinklers or perhaps a smoke control vents that have their own means of activation through temperature dependent links in them. If such a link is broken and devices put into operation in case of sprinklers, that would mean there is a loss of pressure in the sprinkler network, which is picked up by the pump that that makes sure that the pressure is maintained and the valve opens up and from the devolve you get the signal that Sprinkler section is operating and this can be considered as a fire alarm as well. In case of Ventilation devices, if you have smoke vents on your roof, you perhaps can equip them with some Connectors that disconnect when the device opens and you also could say that, okay, this device opened automatically. So there must be a fire to which this device has responded. But In most of the buildings, to detect the fire, to really go through all the complicated array of things, you would rely on the smoke detectors, either optical ones or heat ones, or a multi ones that that we have just discussed. So that's pretty quick one-on-one on how we detect smoke and fire in buildings. Perhaps a topic that needs a much more in-depth discussion because there's a lot of things that go in there. But as long as you're using is smoke detectors that are tested, that were certified, against specific test fires. That's a methodology to assess if, if, for example, a sensor is applicable to detect smaller inquire or a flaming combustion fire or a Solilo's fire, all those fires are defined in the test fires, so called, and those sensors are tested against them. So we know that those sensors are capable of detecting fire. But now imagine the scenario you have a fire in your building, you have some smoke produced, the smoke goes into your sensor, whatever type it is. The sensor detects the change in the environment. And At that point, what does it do? Well, it can either simply issue an alert. That's one way. Perhaps 20, 30 years ago, that's how it would work. A sensor would simply break a threshold of Signal that is monitoring and send an alarm signal. Further, today a lot of those sensors would have internal processing in them, which means they would actually process the data and Assess, based on some Signatures they know, if that is smoke or not. So already you would have a layer of false alarm protection built within the sensor, or the sensor could also be a part of a much bigger integrated system where it's just acting as the measurement point and it is a control panel somewhere in the building that gathers all the data from all of the sensors and and processes it into Information is the fire or not? Based on what the sensor is reading. So it can be from very, very simple like imagine a Temperature sensor with a threshold value. The threshold is passed and the alarm is given to a very complicated, really smoke probing and real-time assessment of the conditions in the building to determine if there is a fire or not. Why we need that? Because false alarms are kind of huge issue in the industry. You don't want to have false alarms. I mean, on one hand you could think that Whenever there's alarm you should evacuate the building and provide safety to its occupants and if it's a false alarm then great, no one was harmed and we have succeeded as fire safety engineers. That's one take you can have. But if you do that few times the user will face consequences like there's so much disruption coming from false alarms that At some point the user might may value their continuity of of operations over fire safety. And if that decision is made then the user will Influence the fire safety systems to reduce the number of false alarms on their own. And if user does that then it's probably gonna end up badly. If anyone is to fiddle with the systems to Improve the resilience against false alarms, it should be fire safety engineers and the manufacturers of the systems. The users usually do not do a great job because they're simply not dedicated For for doing that. So that's why we need to minimize the false alarms as much as we can on our own To give our systems chance to stand against the users. And I'm now a little afraid I'm gonna sound like a broken record after this episode and with user this user that all the fire safety Systems in the building would have worked if not that damn user in the end. But that's true. That's the reality. The human component is usually the weakest link in every Automated system and in fire safety systems it's no different. So where the user comes in, imagine your ancestors have issued the fire alarm signal and there's a device Fire alarm central, we call it in Poland that received that Signal. And what does the central do? It doesn't immediately, like a vacuum, is that start ringing a bell and evacuate the building? It does not do that. It starts executing a predefined series of events and Time delays that are built into the system. First, the control panel must know or must receive an input If the operator is actually present. So if someone has built a 10 minute delay for operators response in the system, but the operator went to eat the lunch and they're not in the room with the control panel for any reason, and the control panel, well, wait 10 minutes for their reaction when they are not present, that's not good right. So the control panel probes if there is a user by issuing a pre-alarm and the user has to respond to that pre-alarm, usually by pressing a button, and that is usually quite a short time, let's say 30-60 seconds, just to confirm the presence of the user at the control panel. The second delay that's usually built into the systems is a delay for confirmation that's a very critical one. It's a time at which the user has to decide whether the fire hazard is real or not in the building, whether the threat of the fire is real in the building or not, and that time can take from one to a few minutes, that the user has to literally walk in or use any other mean to verify that, and within the time the user can either confirm the fire alarm by pressing a button, which means yes, there is a fire, let's go, or the time can pass, and then the fire alarm would simply go into the alarm mode on its own. There's also more things that can happen in this time that shorten this delay, because, again, we don't want to have this delay too long, because this is the time where only the systems know about the fire. The building occupants don't know about the fire, the fire brigade does not know about the fire, the user is notified that perhaps there is a fire, but that's it. No actions are being taken yet, perhaps some pre-actions, but that's later in the episode. The thing that may happen is if another sensor detects a fire, then the system would usually go into full fire alarm mode. The reason for that is that you may have a false alarm from a single sensor. That can happen, yes, but in most of the cases. To have multiple failures at the same time, like multiple sensors going into a false alarm at the exact same time the probability of that is very close to zero. So if you have alarm issued by multiple sensors, it means that there is something happening in the building. It doesn't necessarily guarantee that there's a fire, but there is something happening and the alarm should be issued. So those delay times in the fire alarm system, as I said, they play a critical role in delivering safety in the buildings. And there's one more thing that's not written in the book, that's not a part of the scenario but unfortunately happens more and more often is that the user has also one more thing they can do and that is to manually delete the alarm or manually turn off the alarm. So imagine scenario the fire is detected in the building, the fire alarm probes that the user is there, the user responds and then the fire alarm goes into this mode where the user is supposed to probe for the fire, but the user knows that it will take them 10 minutes to walk to a specific part of the building and they have only three minutes on the fire alarm system, so they can pretty much delete the alarm and go, and then the fire alarm would be reissued after some time again and all the times would restart from zero. Or perhaps even worse. The user could just know that there was an issue with sensor in room 101, and the sensor is always broken and now it goes into alarm and it's just annoying. So it turns it off, then it goes back and turns it off, and goes back and turns it off and after 15 minutes realizes it's not a broken sensor but it's a fire. Really, such scenarios happen in real life more often than you would think Systems being turned off, users fiddling with the systems, users manually overriding the alarms. This is a big issue and of course in the post fire litigation process it's being taken into account that fire safety engineer has done their job. It was user that has broken the system by their actions and regardless, if you want to design good fire safety systems, it means systems that the user would not have to turn off, systems that cater to the needs of the user, that understand the needs of the user, systems that don't go against the user in a way that forces user to turn them off because, yeah, that's a serious hazard in the building. Many of us would not take that into account and I've seen way too many buildings in which this happens. Anyway, for the purpose of the rest of the episode, we have to assume that the user has not broken our system, that the devices have picked up the fire and they've issued the signal somewhere. So now where does it go? In Poland there's a big discussion about should the fire alarm panel steer things or not? Is its role limited to just detecting the fire, or should it also give orders, or should it also trigger operation of other safety systems, like directly in your building? And I think for very simple systems like very small buildings, perhaps that's okay-ish. For complicated systems, that definitely should not be the case. There are specialized devices, specialized control panel systems and specialized integration systems that are the workhorses of running the orchestra of systems in your building. So in large building, where you would have a specialized integration system that gathers the data from all the devices in the building, that understands the state of every device in the building, that is able to dictate the role of every device in the building, send signals, trigger them open, close and so on. It's this system that is going to execute the planned scenario of what should happen from this point. The FireLamp system will give its knowledge about where the fire has acute and by the name of the zone in which the fire was detected, the system, the integration system or the control panels can pick the correct operation scenario, which literally lists them. What should happen from this point with all the devices in the building, literally every single device in the building, for every detection point in the building, would have a known state should it be on, should it be off? What should it do? If you think about building like a massive stadium or a shopping mall or 300 meter tall skyscraper with tens of thousands of devices in them, you could expect this table of operation to be pretty big. And yes, it is Like I've seen tables that, after print with a font size like seven, they would not fit on the wall of the building we were in. That's the size of the matrix that is created to define the role of every single system, every single device in your building, for every possible fire location in the building. A tremendous amount of work, something where it's very easy to make a mistake, but there's multiple people going through that, from the fire engineers who define those groups of devices, what should they do in fire. Through programmers who have to insert it into the system, into compliance engineers who check if the system does exactly what it is supposed to do in every possible fire scenario. A painstakingly hard work at the building and we're talking about the end of construction of the building, where there's always rush, there's always pressure. We need to open the building, we need to deliver it. We cannot risk any more delays in the building process and you are left there with hundreds of scenarios to test and tens of thousands of devices to confirm that they operate like they should. Difficult job, a very rewarding job in the end if you are sure that you have delivered the safety of the system as intended and have captured and eliminated a lot of errors throughout. That it's definitely something that we like to do at ITB. And one more thing that may not be so obvious with such big matrixes is it takes some time to run through it. Depending on how the software is constructed, there can be multiple what if? Loops. It's something that sometimes happens in the programming of those systems and it can actually take seconds or even minutes of time to run through all of the commands and this is an inherent thing to the fire safety system design in your building. And if there are any delays that could come from simply going through the program, then those delays have to be included because of the consequences that we will talk about when we reach ventilation and compartmentation. Anyway, there is a massive matrix that tells you what's gonna happen. First, in order to assure that the safety evacuation can take place, you need to notify the people in the building and you also have to clear the pathway for them releasing any type of control devices that hold doors in your buildings so it's safe to exit. And, by the way, that's one of the reasons for the delays built into the systems. In some cases you don't want to release those immediately for security reasons. In terms of notifying the people, you have multiple choices. You can issue a simple alarm, like a bell or a siren in your building, or you can have a more complicated voice alarm system. With voice alarm systems, you need to subdivide your building again into some areas, zones, whatever you call them where you give a specific message to the people that they are in danger, they should evacuate, and then you can have the rest of the building, let's say, stay safe in the location where there are. Or perhaps you could have a more complicated evacuation plan put in place where you evacuate floor by floor. That's how we do it in skyscrapers. We would evacuate them floor by floor, not everyone at once to overwhelm the staircases, but rather floor by floor. So the flow of people is the most efficient through in the most complicated systems, and that's something that we perhaps will have in the future. More commonly are systems where we monitor where the people are in the building. We count how many people go through those, we know how many people are there on the staircase, we know how many people there were in the building altogether and we can drive the process automatically. We can send people to locations where their path to safety is shortest, where they are not threatened by the fire. We can use dynamic signage to guide them through the building. We can issue those messages on the fly. We can decide on the fly when a particular person or group of people should evacuate to give them the best outcome in the case of the fire. So very interesting development in there. Quite complicated systems, quite challenging systems into execution and, of course, the most difficult part of those systems is the human component, where the people listen to the signals given to them. That's not so obvious, but, yeah, definitely something we can look for in the future, and that's how our evacuation systems will look like in 10 or 20 years, I guess. Okay, so our humans are notified, people are evacuating like they should, the systems know where the fire is, the signal is issued through the network throughout the building and all the devices are entering their operation. So why I've said there could be tens of thousands of devices? What really is there in the building in such numbers? First of all, the basic, fundamental layer of protection is compartmentation. We need to compartmentalize the building so the fire is enclosed within a single compartment and we make sure that the fire does not spread through other parts of the building. Of course you have firewalls and all the passive means that make sure that the fire cannot spread through those walls very easily. But in our buildings we have a ton of openings. You have all the gates that connect bigger parts of the buildings. You have all the ducts that distribute air and help us with climate control in the buildings. You may have a ton of different openings that are needed for some purpose of the building operation that connect different fire compartments. You will have duct arrays that are made for smoke control, that perhaps you have one big fan at the end and this fan can extract smoke from all ends of the building. It's just you have to guide the air from a correct place to that duct so you may want to close all the other pathways in the building. We assume that the fire will be in one space in the building, so it makes sense to make a network of ducts that reach to different parts of the building and connect to one single massive fan to extract. That saves a lot of money. So you would do that. So in all of those locations you would have smoke shutters, fire dampers, things that must close or open depending on the location of the fire, and usually those devices would be in thousands in large building grid. Each of them has their own actuator, electrical device that changes the state, opens or closes the device, and they have to be connected to some sort of logical operator, some sort of control panel that takes a signal from previous control panel that tells them there's a fire, you need to shut down or you need to open up, and this logical signal is changed into electrical signal that powers the motor of that shutter to open or close. It takes a while for those devices to open or close. It usually should be less than 60 seconds, but in some cases it could be longer. Now this is very important, that it takes time to open and close all of them. It's important because, if you imagine, you have an array of ducts that connects into three different compartments of your building and they merge into one Major branch in the middle and this branch connects to the fun you want to extract from branch number two, let's say, and that means branch number one and Branch number three needs to be shut off because you don't want to extract clean air from those locations. You want to extract the air from where the fire is right. That's the purpose of smoke control. So you need to shut down those other branches now. It takes 60 seconds to close them. If you start your fan too soon, you may start extracting air from the other spaces in the building and this flow of air may interfere With the closing of the dampers. So you cannot do that immediately. You have to wait until all the dampers have changed their position and that is only when you can really start extracting air. The same goes for any sort of Makeup air sources. So if we extract large volume of air from a specified compartment and that Compartment is not really connected to the atmosphere with any other openings, you need to provide makeup air into that compartment. If you don't provide the makeup, it means you will create a significant pressure difference between this compartment and the outside, which means there will be strong forces acting on the connecting interfaces doors, windows, gates. If there is a mismatch in time between delivering the makeup air and starting the extraction, you're gonna have a bad time. And, trust me, we've seen that multiple times, especially in car parks where you would have those automated, automated guillotine type gates that would shut the fire compartments in car parks. You start the extraction, you don't have makeup air there yet because the shutters on the external opening, for example, has not opened yet. You quickly create a huge pressure difference between your car park and the exterior, and the weakest point is your gate. And what the gate does it collapses, it opens up and the air finds its way into the car park Definitely not an outcome that you would want as a fire engineer. So it's not just that the things have to happen, they must happen at the very precise, very correct timing. Another example of timing is when you have HVC systems in your building. So of course we need the climate control in our buildings. We need to provide clean, fresh air to the building occupants. Sometimes you would have systems that are fundamental for the processes within the building. Hvc is is a necessity and this means air is transported throughout the building across fire compartments, usually because you would not install HVC for every single fire compartment. It's a building. Until there's a fire you don't care that much if it's a fire compartment or not. Those systems, ductworks that they have, are Potential routes for fire spread or smoke spread throughout your building and you will need to mitigate this hazard. You need to shut down those HVC systems, usually as soon as you can. The other reason is that HVC systems can also interfere with smoke control. By introducing a stream of cold air into a smoke layer, which usually ends in Completely breaking down the smoke control operation. They can interrupt the detection process on its own. So also that that's quite challenging. In general, you don't want HVC operating when you have fire in your building. Now In some buildings it's not a big issue. We would usually shut down HVC as soon as a pre alarm is issued by the fire alarm. So remember when I told you that the detectors detect the fire, that they issue the alarm to the fire alarm central, then there are some delay times built into that process. You can actually take that into their advantage and already shot the HVC the moment the pre alarm is issued, and that's usually quite helpful in some of the buildings. However, in Some cases in many cases actually this is not possible, because Restarting HVC for the building is a procedure that takes hours of time, and in that case you probably would like to delay the HVC shutdown Until the fire is confirmed, which means the operation of the fire systems in your building have to be delayed as well, because you cannot start operating your smoke control when there's another ventilation system operating in the building. So first you need to shut down the HVC, then shut down all the dampers on the HV system to provide compartmentalization, and once that is achieved you can only start your smoke control systems. We're talking about minutes in here. We can have delay in minutes until the smoke control systems are actually triggered and turn on and achieve their full capacity. This this is really challenging because if you think about the evacuation process A set, r set calculations. They are in minutes, literally, and it's not that your systems will be operating there immediately. Those delays must be understood by the fire engineer and must be included in their engineering analysis. It's it's really serious. So now let's assume we are in the good scenario, the optimistic scenario, where the fire was detected, the user has not interrupted it, the signals were issued through the networks of the building and every device has shut down or triggered on For the correct location in the correct time and those devices have not broken each other. This orchestra has worked as the fire engineer has defined it in their fire strategy and we are having a building Baffling against fire, just as the fire engineer has expected it. What does the building need to continue this battle? It's obvious it needs power. You cannot operate those systems if suddenly your power shuts down. And the same is for firefighting operations. You cannot battle fire if there is no water supply to the building. The same for sprinklers you will not succeed if your sprinklers run out of water. So it's not just about triggering everything, it's about making sure that the things operate for the duration of the fire. And here we go into another array of systems, very complicated systems. That is, the power management in the building. So, first of all, when the fire is in there and the firefighters come in, they don't want to Put water on things that are under electricity. It's, it's hazardous. They don't want to do that, and it's kind of understandable. At the same time, you cannot shut down the power to the fire safety systems that operate in your building, because you need them to continuously provide the actions for which they were designed for. So you need to design a system that will shut down the power to everything in your building but the critical fire safety components, and that's that's. There's usually a system in your building to do that. There's usually a switch that firefighter can press and it shuts down everything but the fire systems. And yeah, it's a complicated system. Actually, if you ever seen one, it can be a very, very complicated, very complicated design that tells you which exact cables should be on power, which shouldn't not be powered in case of a fire, where the electricity should go and how to make sure that we don't turn off something critical. The second thing is the continuity of the power supply. So I mean, it's a fire. The fire can, for example, damage the power supply cables. I mean that that's a very possible scenario in any building and in that case, well, you don't want the fire to destroy the fire systems. The fire safety systems have to be resilient against the fire. That's happening. So you need to design a specific power supply To to those systems so that in the case where one line is destroyed, one power source is damaged, then another one can take its role and provide continuous operations for all those systems. But in the between, you need to maintain the correct operation of all of the devices. Like you cannot just turn off the power and Turn it on back after two, three minutes, because you just switch the electricity source. Imagine times where our laptops had very bad batteries and sometimes, after two years of using a laptop, you would reach a point where battery was a nuisance in the laptop and you would have to take it off out of the laptop. Good old times where you could just Remove the battery from your laptop in the second without having to remove the glue from your keyboard that keeps it in place. Ah, that's annoying. Anyway, in that case, when you did not have your battery in your laptop and you wanted to take it into another place and you pulled the plug, you had a bad time because your laptop turned off. This cannot happen in building. This cannot happen in the building automation system that protects against the fire. So we need to design systems that, in a split second between the electricity in the condensers, in the those electrical circuit boards runs out, you have maybe a hundred milliseconds for that. In this time you need to provide Additional power for those devices to continue their operation or at least maintain the logical state they are in. This is done through battery systems within your control panels. This can be done by large battery arrays like UPS systems in some buildings. The most interesting one I saw was a flywheel system, like a giant, few tons of a flywheel that was spinning at very quick rotation and as soon as a power drop was noticed, that Flywheel would start, generate electricity in for your systems and in the meantime had a massive 2 megawatt diesel generator would start and then the diesel generator would take over very, very interesting and robust system. I'm still thinking how lithium-ion battery revolution will influence this part of fire safety design, because we will eventually want to replace those power supply devices with less complicated lithium-ion battery storage systems and this of course introduces new challenges to our buildings that we perhaps are not ready yet, because we will need thousands of kilowatt hours of electricity in those systems, thought for the needs of fire safety. Anyway, another topic for another podcast episode. You need to provide backup power as soon as the power drops. For the water supply, very similar action. You need to have continuous water supply from your city network. You need to have some tanks that store water for the firefighting or for the sprinklers for the pre-designed amount of time and you need to have redundancy in pumps, in the systems that deliver this water into the network of your building so the water reaches the place where it is needed. All of those systems need to be managed, maintained, turned on by your fire automation and monitored throughout the operation so that the operator knows that they are working correctly. So, as you can see, I hope I've delivered the promise from the beginning of the episode. It's not just a simple yeah, there's an alarm, let's go start all the systems. They're operating, we, the building, is battling the fire Now. It's a complicated, multifaceted challenge with so many different types of devices that play an important role, from the detectors throughout integration systems, control panels, out-to-do execution devices, shutters, dampers, fans, pumps, all the systems that have to work together to deliver the safety in a building. And I've also promised you the number one issue that we've seen with this complicated interplay of systems in real buildings. And before I tell you that, let me ask you a question have you ever programmed a fire central? Have you ever built up an operational matrix for a building? Have you ever put lines of code into the fire safety systems to get the correct operation of those systems in the building? And I would assume for 99% of you, perhaps even more, the answer is no. We've never done that. I've never done it myself either. The reason for that is that that's kind of a different job. It's a job of programmers, not job of fire safety engineers. And here where lies the massive issue is again in communication. I've said it so many times in this podcast communication is the number one skill that we need to master as fire safety engineers. Programmers will do exactly what they are told in the documentation of the project and what they are told by the fire safety engineer Exactly that. And they need to be guided very, very precisely to achieve the effects that we want. And I've seen in many buildings that when, for example, the programmer was told for Smoke Zone 1, you need to do this operation. For Smoke Zone 2, you need to do this operation. For Smoke Zone 3, you need to do this operation. They would very much put into the code. But on the building we end up with fire being located in zone number one. Detected systems trigger. Everything starts as it should and after a few minutes the smoke has ventured into Smoke Zone 2. The fire is detected in Smoke Zone 2. What does the system do? It shuts down in zone one and then triggers in zone two. The smoke moves further, zone three. It triggers in zone three, shuts down zone two and suddenly we have systems operating on the end of the building where there's actually no fire. Some smoke has flown there. Where was the error? So the programmer was told to program and they did. But they did not understand the concept that Smoke Zone or Detection Zone is not a sealed space. That smoke may venture throughout the building and it will. It will. It's enough that someone opens doors to staircase and some smoke goes into the staircase with them. That's sometimes enough to trigger smoke control system in a lobby or in the staircase itself. So those programmers are not trained fire safety engineers. They do not have to understand all the nuances of how the buildings work and executing their job. They may program into the system something that is actually dangerous, because the situation I've just described to you is actually life threatening. Imagine you have a building in which the systems operate in a space where there is no fire and they shut down completely in space where the fire is present. This is a horrible outcome and I'm afraid I've seen that happen at least 20 different buildings, really 20 different buildings. I'm talking large buildings, not a small residential buildings. I'm talking the largest buildings and, yeah, it's kind of scary. I don't blame the programmers, because it's, you know, an assumption that's silent, an assumption that every fire expert understands. There's one fire in the building and we battle it. It doesn't. You don't have to tell that to your fellow fire engineer, but when you work with someone who's doing job for us but it's not a trained fire expert, they perhaps need to explicitly be told that to execute the job correctly. So yeah, once again, in this complicated matrix of systems, you have to understand that you will be working not just with fire people. You will be working with other stakeholders electricians, programmers, hydraulics who are not trained fire experts and you need to be able to convey what the systems are for, how should they operate and what should they do in the building in order for those people to give them the best chance to create systems that actually deliver what you have designed them for, and if you don't do that, unfortunately it's not gonna work. So that would be it. I think it is an important episode because of my experience with this. We have done HotSmoog tests, as I said, as a tool in the commissioning process. I think it was 200-ish buildings and my group in definitely way more than 200 buildings. That means we've done easily one and a half thousand of those fires within the last 13 years that we are doing those tests. If you ask me how many of those buildings we have seen issues in this orchestra of fire safety systems that I've described in this episode, unfortunately, in almost all of them Like literally, we had, out of those 200 plus buildings, there was maybe five to 10 buildings that would simply operate correctly, fully, with no issues in them. But I would say in 10 to 20% of the buildings that we've commissioned, the issues were potentially severe, like issues that could lead to fire spread throughout the building, issues that could trap people, issues that could lead to smoke spread into places where it should not be, issues that make the fire safety systems turn off, issues like that, critical issues of the systems. So it's more common for errors in the process, more common issues happening in this process, rather than a simple execution and delivery of a robust, resilient system. I mean you could expect that I've just explained you a whole complex world of fire automation in your building. That's definitely not simple. So there is space for errors there and with fire engineers we have to understand those errors can happen and our job is to help our colleagues, our stakeholders we work with, deliver their buildings with the least amount of them and hopefully deliver safety. So thank you for listening to this fire science fundamentals episode. It was fire engineering or building engineering fundamentals this time, but I hope I have shown you a part of the fire safety world that perhaps was not, that was perhaps not that obvious or not that typical. That was perhaps very interesting to you and maybe you will look at fire automation systems in a different way from today. So, dear friends, thank you for listening to the fire science show. Again, I invite you to join the fire science show community at communityfirecienceshowcom, where there already is 250 professionals gathered. It's starting slow, but I see the community growing and more interactions happening Every week. It should be better. I highly welcome you there and I hope to see you here in the podcast next Wednesday. Thank you very much. Bye.