 Hi everyone. We are going to wait for a few minutes for people to join. And we will start the webinar at probably 1202. Thanks so much for being here. Once again, we're just going to wait like another minute for everyone to join then we'll get started. So thanks for being here and hang tight. All right, I'm doing it. We're going to start. Okay, welcome everybody to this bonus New Mexico's Margaret Center summer webinar. It's 100% carbon free energy system can we still keep the lights on with Dr. John all the way from Clarkson University in New York. I'm Brittany Van Dwork, the communication and outreach special specialist for New Mexico, established program to stimulate competitive research. New Mexico EPSCORE EPSCORE is a nationwide program funded by the National Science Foundation. And I'll be your host for today's webinar, along with my partner in crime ISIS Serna, our website administrator who will be working behind the scenes to make it all flow smoothly. So, a few housekeeping things before I begin. I want to let you know that if you have questions at any points, please type them in the Q&A box and ISIS will politely interrupt Dr. John and read them out loud. I also want to take a hot second to tell you about the epic webinar training lineup we have for you this fall. In August, we kick off with Dustin Allen systems and network analysts at the New Mexico State Office, who will be presenting Python fundamentals with data analysis and visualization on the 25th from noon to one. And if you ever wanted to learn Python, Python now is the time. Then in September, we will hear from about research from Dr. Juan from MSU School of Engineering, and he is also a new New Mexico Smart Grid Center faculty faculty higher. Finally, in October, we will learn about the research of Dr. Xiao at an assistant professor at New Mexico, New Mexico Tech Department of Electrical Engineering and also a new New Mexico Smart Grid Center higher. Registration info can be found on our website, as always. Okay. Now that's out of the way. I'd like to introduce our presenter for today. Dr. John, we, who was recommended by the aforementioned Dr. Juan from MSU. Dr. John is an assistant professor in the ECE department at Clarkston University. Prior to joining Clarkston in 2020 he was a power systems engineer at GE Global Research Center in New York. Dr. John led a GE team to investigate long duration storage for renewable grid integration and participated in design of GE products, including distributed energy resource management system, like the large power transformers and converter based distributed generation. Dr. John earned his PhD from Washington State University in 2016, and his research interests include renewable integration and energy digitalization. Thank you so much for being here, Dr. John, please begin whenever you are ready. All right, thanks so much for the introduction here. Let me share my screen. Can you see my screen okay. All right, thank you for the introduction again. Welcome to this talk here. I just wanted to share some of my experience of when we are moving towards the 100% covering free energy system, can we still keep the lights on? This is a very interesting to me that the reason is, now we are more and more dependent on electricity and keeping the lights on is almost the number one priority for our daily life. As you cannot imagine, we have the lights out. The experience I had yesterday that we have the storm here yesterday and then the electricity supply to the water house was out, and then basically I was, I don't have water supply for four hours. That drove me crazy because I cannot cook and cannot take a shower. So that just shows how important the electricity to our daily life. So, given that background, I just want to share a little bit of my view on this topic and hope that I will give you some information, you can get something here. Before I talk about the, how to keep the lights on, I just want to quickly describe my background here. I've already introduced my experience here, just want to give you a little bit more details here. So basically I got my PhD from Washington State University and then my whole PhD program was about outage for both the transmission grid and also distribution grid. And also I have been working on the GE distribution management system for the smart city taskbar before I graduate. So because of that, and then I very naturally and then I transferred, you know, to GE to continue work on that direction. But surprisingly, after I went to GE, you know, all my focus was on renewable integration and also the product design to improve the greater resiliency. The resiliency is almost equivalent to, to say, we need to keep the lights on under the extreme events. So the resiliency is, is mainly to keep the lights on. So in that part, and then we continue to work on some of the technology related to outage. And in 2020, in 2020, I came to Clarkson, I have been focused on the data analytics. So basically now we have more and more data and how to use the data for the decision making power system. Seems like you guys will have the next seminar on the Python learning for the data and analytics and visualization, I would say, I wish I could be the student here so I can have the chance to learn, but unfortunately I cannot but I hope that you can, you can, you can learn that and then you can, you can see that what we are doing here in the power grid also quite rely on the data analytics and the visualization. And then the other part is, I also focus on the grid operation under uncertainty. That part is more about driven by the renewability rated into the grid. Your, your, your renewables very uncertain means you have a lot of variability, your forecasting has a lot of uncertainty. How do you operate your grid? Consider uncertainty. That's another direction that I'm focusing on. But definitely I'm also considering, I'm also continuing the work of outage related work there in the storage related work that have what I have been working on. Since day one, I was in the PhD program. So basically that's my, some of some of my experience. I also wanted to briefly introduce where I'm from and the power engineering program at my university, which is Clarkson University. The top left top figure is the where we are located. If you look at the Clarkson is in New York State and indeed people always questioning is in New York City or is where in New York. And I have been asked, I asked this question a lot of times, and indeed the Clarkson is in the upstate New York. Indeed is very close to the border between the US and Canada. People from the Clarkson usually go to Ottawa or Montreal, instead of go to New York City. If you look at the distance is very obvious. And it takes us about one and a half hours to go to Ottawa. It takes about two hours to go to Montreal, but it takes us about six hours to go to New York City. So this is where Clarkson it is. This is our canvas in the summer. We have a center we established a power system engineering center in 2018. Currently, the core members we have, we have five faculties in the purely in power system, pure power engineering. We also have a coordinator because the center is founded by basically by the Clarkson in collaboration with the industry. We have quite a lot of the members basically we serve the industry in the northeast part of the country. And it has a long history of power engineering. It's a strength for Clarkson and we trying to continue that route and want to use in our expertise and experience to build a great of the future. That's basically the where we are. And my, my experience. So now let's jump into some of the information regarding the generation max in Mexico, because we, when we try to hit the target of 100% cover free. We also need to know where we are. So let's look at the data from the energy information administration is a US agent. And let's turn me now in New Mexico. The energy, the generation capacity conversation, it looks like this. This is data maybe is a little bit dated. The data seems that it's from 2019. But definitely that give us a good reflect of where we are in New Mexico for the general electricity. Roughly 30% of capacities from coal, it's cool. And 37% is natural gas and the wind accounts for 23% solar 80%. This to me that New Mexico is pretty dry, and then the hydro is only less than 1%. This is the base on the generation capacity. And then if you look at the electricity, electricity basically is energy from energy perspective, the coal that generate 42% of electricity from New York for New Mexico state. And then natural gas generate about 34% of the electricity in New Mexico state, the wind and the solar, the capacity why they have about 31%. But in terms of electricity, they account for 24% of electricity for New Mexico state. Very, it looks like we have a lot of work to do to really transform the grid from the where we are to 100% carbon free. One of the definitely the one of the thing we need to do is that get rid of the coal coal power the power plants there. You can look at that the capacity why the 30% the energy why the 42% that will lead to dramatic change of the energy space. People may also say the natural gas is also generating the carbon dioxide. So, I look at the data seems like in currently in the US, about 62% of the natural gas, they are not equipped with the carbon capture technologies there. In other words, during the transition to 100% carbon free energy system and other work we need to do is to run it, either you retire the gas power plants, or you need to invest into the carbon capture to make sure that natural gas will be carbon free. With all of that, the need of the technology or the investment or the change of the generation max. It's just another angle to show that we really really need a lot of engineering for the great decolonization in New Mexico. That has been also reflected by the goal. The goal for the Mexico, it, I think the legislators set up the goal that by 2030 50% will be the electricity will be front of renewable and by 2045. And they should hit the target 100%. If you look at the currently the renewable compensation in the US. That's where we are. In the next few decades. This is where we will go to. Just one thing I want to point out here is, um, here is a percentage in other words, if you look at that percentage now the hydro is about 34% in the renewable. Excuse me, we have we have a question from Anthony Franklin, why no nuclear. You mean the nuclear for for here for the for the renewable or the nuclear for the generation max. Anthony, both. Okay, so this data assumes the new Mexico does not have a nuclear or nuclear is very small. That's what I, that's the front of data from the EIA data there. He says he sees it now. And then that's a very good question. Indeed, I like that question. Why not nuclear. Indeed, the currently nuclear is covering neutral. You're right. And currently I think in the US we have all 10% of generation electricity from nuclear. The nuclear has the challenge is for the largest scale nuclear and the flexibility is very so basically nuclear is not a flexible. Unfortunately in the traditionally we use a nuclear to serve the base load, and also the waste of the nuclear power planes. It's always a concern. So, for the largest scale nuclear power planes, basically is not our favorite from the policy perspective. And also from the market perspective I was saying, it's because the nuclear is serving a base load. In other words, the nuclear should be running constantly at an almost constant output. But because the generation system is transforming to high penetration renewable driven dominated energy system. The system will need a lot of flexibility. But the flexibility cannot be supplied by the nuclear that created the challenge to maintain the secure operation of the power grid with high renewables. Now, in the nuclear space. A lot of momentum is trying to design a smaller nuclear is more flexible. So the smaller the smaller scale the nuclear technology, I will say, the technology now is still in the development phase has not really changed the game yet. But I will envision if the small scale nuclear power plants, the new technology will come into the become mature coming to the market. The renewable composition will be somehow be changed by that. But the condition is, we have to have the mature technology and with a lot of flexibility to really accommodate the variability and uncertainty of renewable. I hope that answer your question. It did. He said thank you. No more questions right now. Oh, I forgot to wear it where I was but anyway, I will. Um, so if you look at the on the renewable profound here. Um, so in the next few decades solar and the wind will continue to ramp up. Hydro. I will say hydro still considered a very key component for the renewable future. And indeed in New York state, we have about 24% of electricity is from hydro. And we are the New York power authority thinks hydro facilities will be a key enabler for New York to hear the 100% of carbon free electricity grid. So if you look at the, um, this is the founder of the, the employment. Um, this is the data is in 2019. Um, in the electric power generation employment, the employment that increased by 4.8%. The transmission the distribution storage sector, the employment increase by 3.5%. This is just another angle to show that. When we are transforming the power grid, and the demand for workforce is definitely increasing. That makes a strong demand of power engineers or engineers in the energy space. And then luckily I saw, I dig into the literatures. I saw one statistics for I trouble. So this is the I trouble you transaction on power effort test and assistant basically that's the top one journal in the I trouble you on power engineering. And I started in New Mexico state. The name is here. Definitely. This is a rigorous program is the fundamental to develop the workforce on board. The New Mexico state or the country to really hit the target 50% renewable by 2030 and 100% by 2045. When the power grid is transformed is transformed into, you know, renewable dominated power grid. Another concern is a blackouts. That has been the effects of the blackout has been shown by the Texas blackout, you know this year in the winter. But, generally speaking, because of a blackout it costs the US economy about more than 100 billion dollars per year because of the outage. And also they have some statistics shows by each state. What's the annual business losses from the grid of programs are definitely different new max New Mexico is, I will say is less impact. More impact by the outages or greater problems. But if you look at the other state like California, or Texas, or Florida, Illinois, or New York. We are highly impacted by outages or greater problems. Today is a hurricane Sandy in 2011. So if you look at her going Sandy in 2011, the New Jersey, the whole the new North Eastern part of the country has been had a pretty hard. And a lot of people lost the electricity supply by that. And then the outage last for quite a few days that has a very big impact on the energy policy for all the states here in the most eastern part. So the electricity I usually say electricity is usually taken for granted until until we are experiencing blackouts. So blackouts definitely as a very, very important topic. When we are transforming the grid into the 100% renewable. Let me just quickly review what happened in the Texas blackout from there we can see why we have the blackout. What should we should do to really keep the lights on. In February 24, February 14, 2021 to February 19, another time period, the freezing cold weather had the Texas region, including New Mexico, if you look at the map here. I'm pretty sure that maybe, you know, some of your experience that extreme cold weather in the winter. Um, here is just a recap of what happened there. So on February 14, the aircraft is assistant operator in Texas already seen that they already project that because the cold weather. They may experience the shortage of reserve. But they didn't project that there are so many generators tripled offline, including wind. So this is what happened in the early morning of February 15. And then, so this is the 1212 AM. They saw that the reserve is less than three gigawatt and then they say we need to do the energy conservation. But unfortunately, even you do the energy conservation. You can still not, you still cannot maintain the reserve margin there because the cold weather tripped off the guest parking. So they had a 120 AM. They issued emergency operation level three that's the highest level. Basically, they said, there is no way for me to do anything to keep the lights on. So they had to do the load shedding. So in that time, they do the rotating outage for almost 11 gigawatt load 11 gigawatt load. That's a lot of customers. I think the whole New Mexico, the peak load the properties around is about it's about this number. You basically cut of the power supply to hold in Mexico state, and then doing the road, doing the road in black outage. So if you look at the, the generation on the outage due to the extreme cold weather there. So this is the published by the air cut is a system operator to for taxes. And that the scale here is giga giga it's a huge number. Um, so the for the 15 the 16 around the 52 gigawatt generation trip the offline, because the extreme cold weather. And then if you look at the composition, the majority leading factor is natural gas and also wind. I will say the wind definitely play a role here, because the cold weather, not a leader to the running blackout. If you look at some of the data here in Texas, the peak demand is 75 gigawatt roughly. And then the whole the total generation capacity they have is about 107 gigawatt. And then they have a generation outage about 52 gigawatt. If you do a single map here, using the 107 gigawatt minus 52 gigawatt, then you get about 55 gigawatt. So 555 gigawatt generation, definitely is way lower language demand. Right. So, that's one of the key feature in power system, we almost need to keep the balance of between the generation and load to maintain the frequency. In other words, your generation is way short than the demand. Then they have to do the load sharing. So, in that in February 15, February 16, the shed about 20 gigawatt load to maintain the reliable operation of the grid. If you look at the root cause of the blackout is the insufficient generation to meet the demand due to extremely cold weather. But if you go deeper into that, if you go to read the report from the air cut, it's because the cold weather makes the gas delivery system not working. So your gas cannot be delivered to the gas fired power plants and the wind turbine does not have the as the weatherization technologies so the freeze out, they cannot generate electricity anymore. So, extremely events or extremely cold weather definitely lead to the running blackout in Texas. 20 gigawatt a lot of shedding is a lot. That's basically bubble the electricity that they made in whole New Mexico state. Then we look at the 2020 the California running blackout. The root cause is is is totally opposite. In 2020 in the summer on the extremely heat wave. Had the western coast, and then including the California. So California experienced the one, one in 30 years weather events, and then the coming change in do the event hit away really extended across the western United States. Indeed, New Mexico is also impacted by the heat away. So for the California if you look at the load curve. So this is the load curve. And then you look at the peak here definitely the load is larger than a typical year. Because the high demand is from the heat away in the, in the meantime, the heat away also. The heat wave also make the thermal generators, that's efficient. We know that thermal generators, they depend on the ambient temperature, the efficiency and also the capacity of the next limitation the generation limitation will be impacted by the heat wave. And so for speaking the thermal generators, they can their limitation, a megawatt limitation will be larger in winter, but it's less in the summer because the higher ambient temperature, and also because the heat wave, then you have the smokes, the solar generation is decreased. So the California experienced the drought, they don't have sufficient hydro generation. So it's a lot of a factor lead to the rolling blackout in Texas. If we summarize it is, is the imbalance between the generation and the load, the heat wave increase the demand, and also decreasing the generation, not yet to the running blackout in Texas. If we summarize the two running blackouts in Texas in California, you can see that on in Texas they have 52 gigawatt generation is offline. The California solar generation decrease gas turbine decrease load increase for both scenarios, and then the impact is a lot of four point four point five million in Texas houses are lost the power, and in California roughly one gig war load. The power shedding are was executed, and then the impact the Texas one is up to four days outage for some of the customers, but the California on the outage was less. So, but people will say, you know we pay a lot of electricity bills, we will make sure that system has sufficient generation to meet the demand. That is should be done should be playing in the design of the power grid. Why we have not prevent this kind of blackouts, or say, if we have a blackout, what should we do. I have to claim that the engineering practice for the energy system are really really challenging by the extreme events to keep the lights on. In other words, the engineering practice in the power industry has not considered all the scenarios that are of the energy system or the scenario including the extreme events or cascading failures of a power grid. So that's that's the practice. Indeed, we can say we can always always improve the engineering products. That's right. And how to improve the engineering products and also considering the cost. That's another thing we need to keep in mind. How to design and operate the grid in a more cost effective way to keep the lights on while hitting the target 100% carbon free. That is our goal. If you look at the historically the blackouts in the US, I just listed some are very typical, I will say signature blackouts in the US history. The most famous one, I will say one of the most famous one is was 1965. That's the northeastern blackout. Indeed, New York was heavily impacted by this. That's almost four or five decades ago. 30 million customers are affected. Indeed, because of this blackout that lead to the installation of a remote terminal unit. And the substation measurement sensors, and also energy management system, energy management system and for the trans a bulk power system management system in the control room. So what is because of blackout people realize that we need to have more visibility into the grid operation. So that's one of the very signature blackout in the US history. Now we look at the 2003 Northeastern blackout. That blackout costs about 45 million customer lost electricity. That drives to further investment into the grid. People of the industries want to have more visibility into the grid. What causes it? They put a lot of investment into the basic management unit. That is about two decades ago. Now we look at the 2011-2012. In that time period, the US has been impacted by a lot of hurricanes, extreme events like hurricanes and thunderstorms like that. About that impact, so the, I will say, this blackout or related to transmission grid. But in roughly one decade ago, that extremely events have impact the distribution grid. About 4.2 million people are affected. One decade ago, people bring the resiliency into our table and say, we really, really need to design a resilient power grid. And then the 2020 and the 21 now we have the control blackouts about 4.2 million customer affected. And what will come out from this we need to see. But I'm pretty sure the renewable. Somehow, when we are doing the transition of the power grid into 100% renewable, we need to keep the blackouts into our mind. And how can we keep the lights on when we do the energy transformation. So I always call the blackouts, it's a wake up call for change of grid, especially the change of the policy change of a new standard change of the engineering practice for us. If you look at the root cause of blackout, there are so many reasons to cause the blackout are in the taxes in California is the extreme cold weather extremely extreme the hot weather. If we categorize that into the top 10 root cause of blackouts there are outages there. The number one is a natural disaster, including the extreme cold or hot weather, as we talk about the second root cause is the motor vehicle accident. So basically someone hit the pole and knocked down the power line that caused the black cause the oddage. And the worst so the third one is the equivalent failure. Basically, I need to say that in the US, the power grid was was designed to build was the majority of property was in 1960s or 70s. In that three decades. So the transmission line has been there for 50 years, the power plants. I will say power plant property is fairly newer, but the transformers and transmission line has been there for a few decades. The average transmission transformers in the transmission grid. The average age is about 27 years. So you just imagine how old they are. Basically, your father designed the power grid. Now you are still using the same power grid, your father design. But that all of that, no matter is the natural disasters. It's the motor vehicle accidents or equivalent failures or it's, it's a falling trees, no matter what. Usually the system operators is sitting at the control room. They don't know what is going on there. They don't understand they don't know if you are experienced experiencing outage until you notify the system operator, either by calling them say, I have an outage. They need to take some action to fix it. Or some, some sensors at your house can detect there's an outage and send a notification to the control room to notify the system operator that you have a power outage. So no matter what, what it really costs, it is the system, system operators relies on the technology to estimate where the outage is, where the problem is. For example, if there is a file, or if there is a, the poll locked down, they need to predict what where the problem is. So they can dispatch the crew to patrol and to come here on site to fix it. So we call that as outage management. Um, before I go on, I just wanted to see, as is, do you see any questions? No questions at this time. All right. Feel free to ask some questions and I will be more than happy to answer any question there if there is. So here I just wanted, you know, going to some of the technology that I have been working on. For example, the first part is outage management of the electrical power distribution systems. When talking about the distribution grade outage management, basically, when you have a power outage, either you call you pick up your phone call to, to notify the system operator saying you have outage, or the smart meter at your house will report outage. So we need to talk about the current in the US, the energy meters we're using. So energy meter is nothing new, but depends on the technology, we have three categories. One is the traditional matter we call the standard meter. So 2000, the automation, we're trying to increase the automation into the power grid. And then roughly in 2000, we started in the US, we started to deploy the AMR, deploy the automatic meter reading meter. But that is not a smart meter. In the decade, we have seen a lot of installation smart meters compared to different technologies. Basically, the standard meter and the utility will need to come despite some, some people here to come into your backyard to read the meters, how much energy you consume every month. That's manually, basically we call it a manual meter. The AMR meter basically is called automatic meter reading. What you do is they have the one way communication. So instead of coming to your backyard, they just need to drive a car along the road and to pin your meter and to read the energy consumption or detect if there's any issue or outage in your house. Now the smart meter is, we call it a smart because you have the two way communication, it has more capability. They report the data energy consumption every 15 minutes or every hour. Beyond that, if your house experience an outage, you can send a notification to the system operator automatically. I look it up the profile in New Mexico State. Currently, the majority of your customer in New Mexico State is still this standard meter. This is about the count for about 62% and then about 12% of the customers is using the smart meters. This number is for the US roughly about now I think about 100 million smart meters has been started in the US. So, because of the smart meters, you can see that. The system operators has the visibility into the residential or commercial. So basically in the customers energy consumption, because the data. Originally, the standard meter can give you the manner give you the data one data point per month. Now the smart meter give the data every 15 minutes. So, the meter, I think that the volume of the data is, is hundreds, tens of hundreds, it tends to 100 times the data volume. So that data give you more information. So my research on this topic is trying to use the smart meter data to help the audience management. In other words, we use the smart meter data to help the system operators to infer where the problem is, say, the, the power to knock it down, or the squirrel like, you know, jump into the, the, the lines there to create a shortage, whatever. And then if you look at the, this is the distribution grid. This is the schematics. So, this is a substation, you know, the substation from the feeders you go to the, the fuses for protection, then you go to the streaming transformers, they serve the house. If there's an outage, for example, if there is a file here, then you trip open the recloser. Then all the customers downstream will return outage, then the smart meters will report outage, send a notification to the system operator, the system operator using that notification, try to infer where the fault is, and what happened there. So the question is that how do you use the meter data to infer the outage scenario. Also, I need to note that, in addition to the smart meters we have, along the feeder level, we also have remote file indicators, we also have the automatic reclosers like that. So the feeder level sensor together with the smart meter, how did you use the data to for the decision making for the system operator. If you look at the outage management issue, you don't know what happened in the scenario, you try to infer the most credible outage scenario supported by the evidence, the meter data there. And you also need to constrain that the physical rules, the physical rules means basically considering the physical property of your system. For example, your protection is coordinated basically when there is a fault downstream of the protection, your protection is expected to activate. You also have the other constraint saying the fault indicators send the notifications only when the fault indicator is upstream of the fault location like that. And then you give all the constraints, you try to infer using the data from the meters, either the smart meters or the feeder level meters, you try to infer which outage scenario of the fault location and also the activated protection is the most credible. But indeed, you don't know what happened there, you don't know the ground truth. What you can do is a purity data driven or evidence driven, try to infer the scenarios there. So in the optimization perspective, you put an objective function, you put a lot of constraints there. You try to solve the optimization, but indeed, because you don't know how many fault out there, if any meter is failed or any meter malfunctioned. In other words, you don't know the uncertainty of the meter data. And then we propose using the hypothesis testing. In other words, I don't know what it or how many issues are there in the system. I can put a hypothesis assume how many issues there for each of the assumed scenario will run the optimization, try to improve the efficiency of the analytical model. The leading model is the challenge is that we, a lot of non-linearity and also computational complexity and the local optimality. And then we propose using the multiple hypothesis methodologies. So you generate the hypothesis, you collect the all the evidence from the smart meters or feeder level meter data. Together, you design the optimization model, consider all the constraints, your remote file indicators should be upstream of the file location, your smart meter audio reports should be downstream of the activated protections there, like that. And then you run the optimization model. And for each of the hypothesis, you determine you calculate the credibility of this hypothesis. In other words, you, you try to rank the credibility for each hypothesis, then using the credibility, you infer which audio scenario is the most credible. And by the evidence you connected from the smart meters or other sensor sensors along the feeder. Now we tested this, the technology using the real world feeders. This is the from the Washington state. And this is the simplified the schematics of that and then we have the scenarios there. We, we generate eight hypothesis for each above this, we run the optimization, we determine the which device protection activated, where the fault is, we calculate how many small meters are aligned with the determined audio scenarios, and then the credibility. We can see that for this one for the other half either hypothesis, we can see that different hypothesis will have it are have different credibility, credibility. We rank the credibility, we find the most credible one, using that hypothesis results to infer the audience scenarios. So basically we infer what happened in the audience, using the data. So in line with what we do in the decision graph, we have the distribution management system will use the data will leverage the data, we try to infer what happened, we call it data analytics in the power system. So this is the distribution great for transmission. For transmission, though, I will say is it's very similar, but the transmission is way more complex than the distribution. The transmission transmission basically the transmission line in the substations in the substation, because the substation automation. We have a lot of sensors in stock into the substation that include the digital relays, the phaser management units, and then other ideas, like all the meters there. And then for transmission, the protection principle is totally different. The distribution grid is way more complex. For each of the components in power system, we have a dedicated protection scheme to protect the component. The reason is one transmission transformers will cost you quite a number of million dollars. So if there is any events, I say, there is a file within the transformer, you want to trip open a circuit breaker to isolate a transformer transformers as soon as possible. Usually, that is done within one or two seconds. So that we have very stuff, you know, dedicated the protection to protect a grid asset there. So usually for transmission grid, for the grid asset, we have the main protection, the primary backup protection, secondary backup protection, breaker failure protection. All the purpose is trying to isolate the file as soon as possible and considering the abnormality and security of your protection schemes. In the substation, you have the PMU, you have other device IEDs, digital relays. You also have the sequential unit recorder that basically to try to record what happened in the substation. Did your digital relay trip, did your circuit breaker open, did your PMU to record the data like that. They all will record in the sequential unit recorder. And the problem is how the front assistant operator, we don't have engineers in every substation. The system operators in the control room, they rely on the alarms or the data from the sensors to infer what happened, which transmission line it followed it, which protections it's tripped open like that. They use the data, using the data analytics technology plus the domain knowledge of power engineering, try to infer what happened there. We call the event diagnosis for transmission. So basically you have the grid data, you have sensor data, you're considering all the domain knowledge of the protection and the system there. You try to infer which component is followed it. Where is the failure of the circuit breaker, if there is any failure or malfunction will arise, or if there is a missing or incorrect alarms like that. So we call that event diagnosis. So here's just one example. A file that happens in my three and then this circuit breaker keep tripping open, but this circuit breaker did not. Instead of the second secondary backup protection at location, circuit breaker one circuit breaker two circuit breaker 15 circuit breaker 12, they took the open to answer the file. Then this is the event, this is alarm we get. And then using the analytical model there, we infer what happened there, at what time the fault occurs, and which component fault it, which protection relay tripped open or failure. And if there are any other time type issues with circuit breaker like that. So, what are we done for here is we propose the analytical model to handle the complex scenarios with the abnormalities. But then, I will, I will go a little bit quick here, as what if the cascading events leading to a system of white black house, in other words, what if your abnormality or files has not been isolated by the protection, or the issues propagate into the graph in the graph. That is what exactly happened in 2003 was spread blackout. So when we have a widespread blackout. What do we need to do we call the power system restoration, what is a power system restoration basically, you have some dedicated a black star units, you use net a black star unit, you try to crank the network, and also the generators, you try to restore bring back the transmission line, and also the generators back into normal operation. Step by step. So we call the generator restoration system restoration that loaded restoration if you look at the pictures basically, you are using one of the dedicated generation unit, we call it a black star unit, you try to crank the network step by step, we call the restoration there. Indeed, that effort was from the 2003 that the widespread blackout and then every was put some initiative to design the software there. It takes a long time to really design, develop the tools there. Indeed, one of the tool is divided by, I will I want to say to purely by me but started from me. Now, they tested the work with the technologies across the world for for this technology. I do have a small video here to show it's a small, a very quick clip to show that the tool here. Restoring power to the electric grid after a total shutdown. This ability to black start the grid is something all grid operators plan and practice for and it's one capability they hope to never use. A black start involves using designated power plants known as black start units that can start without the help of the grid. As these units return to service grid operators methodically connect electric load and other generators to restore the system through prescribed steps. While the concept of a black start has generated. Sorry. In a discussion following the extreme winter weather impacts in Texas, restoring electric service in the state did not require a black start. That's because the grid is designed to withstand disturbances without leading to a blackout. Grid operators must anticipate and mitigate a number of scenarios from severe weather to natural disasters to cyber attacks to an electromagnetic pulse. The Electric Power Research Institute works with utilities around the world to harden their systems plan to mitigate the impacts of extreme events and expedite power restoration. While some estimated that restoring the Texas system from black start could have taken months, the black start of a de-energized but functional energy system would have taken a matter of hours or at most a few days. Epri's approach to black start planning consists of two primary components. Epri's optimal black start capability tool finds the best location of black start resources, those that power on first, as well as the sequencing across the grid to restore priority loads and non-black start generators. This identifies the ideal black start strategy, assuming all resources are available. So, given the time, I will not stop sharing the screen, sorry. Given the time, I will not go to the details of the other videos here. So this is the basic tool and also we tested the tool using the energy systems. There's some results there. Given the time, I will not go to too much details here. But then when we are looking about the transformation of the grid into the future and then we are really have the challenges here. In addition to the challenges that we have been talking about in the last decade, no inertia from the renewable variability and uncertainty from the renewables. I think the most challenging one is engineering practice and the controllers in the system now currently we have. They are all based on the conventional generator resources. So that created, I will say, the most challenging part for the transformation of the grid. Now, this is some of the, yeah, sorry. John, it looks like we actually have a question from Anthony. What is, and so I was going to interject because you have two minutes left, maybe you can use the two minutes to answer questions. Did that work for you? Yes. Okay. This has been a fantastic presentation. Thank you so much. Anthony's question is, what are the biggest obstacles currently facing you and your work and what are some of the future challenges you're going to pay? Right, I think that's a fantastic question. I think the challenge you want is how do you understand the value of the data now from the grid. Understand the value. And then the second is that how do you design the technology that are based on the data combined with the domain knowledge to really unpack the value of the data. For me, one of the challenges is that I sometimes I cannot get really the data that I want for my study. But from the technology perspective, the challenge is how did you design the technology to really take the challenge of it faced by the grid? Awesome. What a succinct and on point answer. Thank you. This has been an absolutely fantastic presentation. There are not enough people in academia that present like you do. So thank you, thank you for being here. I'm going to quick close this out since I don't see another question and take over the screen so I can show this. Yes. Thank you. Thank you again, Dr. We've got someone in the, okay, cool. Thank you Dr. John once again for being so generous with your time. This is an absolutely fascinating topic and when that's, as you know, at the core of our projects research mission so this presentation was quite a treat for us and and for those of us who are non experts in the field it was fantastic. Before we sign off I just want to thank my partner in crime and Dr. one for suggesting Dr. John as a webinar speaker because it's fantastic. So thank you. Thank you for presenting. Thank you for having me here and I thank you for all the students of what 10 or faculties attend the seminar. If you have any questions feel free to reach out and then you can easily look at my information website and I will be more than happy to keep the conversation going there and keeping in touch with you. This has been a great day. Thank you so much. And don't forget everybody to join us again in August for Python fundamentals data analysis and visualization with Dustin Allen. Until next time, have a great, great afternoon everybody.