 And today we have the pleasure of having Professor Jaquie Wong. He is from the Southern Methodist University. He was a past editor-in-chief of the Ikebuli Convection Smart Grid and a distinguished lecturer. And his research focuses on smart grid, power system operation in the control, and grid resolution, diversity. Thank you very much, Jaquie, and we will be close. Thank you. OK, thank you so much for inviting me here. Again, my name is Jaquie Wong. I work at the Southern Methodist University in Dallas. And it's my second time here. So last time I was here, it was like eight years ago. So it's good to be back. Today, my presentation is about the integration of hydrogen into energy systems. And I look at three different aspects of this integration. One is the reliability assessment. The other is optimal operation and planning. So this is the outline of my presentation. So I'm going to give you a general background about hydrogen integration first. Now I'm going to address this reliability assessment problem with the integrated energy system. Considering hydrogen. We'll look at this optimal operation and the optimal planning of an integrated energy system with hydrogen. Then I will conclude with some remarks. So this is the general background of hydrogen energy systems. And you can see this is from this report from IEA. It probably is titled Hydrogen at Scale, H2 at Scale. So basically, it lays out this roadmap, this future about this hydrogen-based integrated energy system. So you can see that we have very grand goals. And they project that the total hydrogen demand is going to increase to 528 million tons in 2050. You can see that's a big number. And also, we envision that hydrogen is going to be used in a lot of sectors in our industries now. The renewable, RE is renewable energy based on water electrolysis is going to play a major role. So they're going to help mitigate the renewable energy curtailment, which we are having now. And also, it's going to help with the realization of that zero emissions in 2050. So that's the goal. And also, the hydrogen fractions in global natural gas system will increase to 15% in 2030. I'm going to talk about that. The way that we use hydrogen, one of the ways is that we can blend that hydrogen into our natural gas system. So you have a mix of hydrogen and natural gas in our gas pipelines. So you can burn that, you can use that. And we also have this goal, 62% of hydrogen from electrolysis, 30% of hydrogen is going to be used for electricity production and fuel by 2050. So you can see, this is the goal. Then the issue that how you can achieve that goal. As you can see from this figure here, hydrogen is used that hydrogen basically is a coupling connector between the electric power infrastructure and also the gas system. And you can see that because hydrogen can be generated from gas, can be generated from electricity, and hydrogen can be also used to generate the gas or generate electricity. It's like a two-way energy flow. So that's the legal between the grid side and also the gas system side. So, and this is the roadmap that our PVO people are talking about. So in the near term, the hydrogen is mostly produced using this steam-methane reforming process, SMR. So basically you'll use gas to generate hydrogen. And the second one is a grid that electrically-based electrolysis. So we have power from the grid. We have this electrolyzer that you can, that you can spit the water into hydrogen. So, and also in terms of transmission, transportation we can use is a gaseous tube trailer. And also we have this small-scale hydrogen pipes. We don't have a long-distance hydrogen transportation pipes now. But in one of the conferences that I went to this year, that Europe is building this intercontinental hydrogen transmission pipe, very large scale product they are doing. So that can be the future tool. And also hydrogen can be blended into natural gas system and to be used now. So mostly the industry is using hydrogen for now. But over the long term, people have those kind of futures. So hydrogen can be produced through this steam-methane reforming with carbon storage. So because when you use gas to generate hydrogen, you generate CO2. So then you can use CCS to capture those CO2 and store that. And also in the future, we are mostly talking about the blue hydrogen. Hydrogen has different colors. You have blue, you have gray, you have different colors. Different color means that they are generated by different sources, by different ways. So blue hydrogen mostly is a way of generating hydrogen. So basically you use renewable energy to generate hydrogen. So in the entire process, there's not much CO2 involved. So this color are blue hydrogen, renewable energy based electrolysis. And high temperature electrolysis is also another way of doing that. And so also we can use liquid shock to transmit hydrogen and build this large scale like Europe. Those hydrogen pipelines. And hydrogen can be consumed by industry, by transportation, by building, by power generation. OK, so we have those ground goals. So we have that road map. We have that mission, how you can get there. So that's the reason that we are trying to address here. So along the road, you're going to have different technical barriers, different technical challenges that you're going to encounter. So I'm going to talk about several of the topics that we can think of now. And then we are trying to solve that. So one of the topics is the dynamic energy flow of integrated energy system due to issues associated with computational accuracy and efficiency. Because when I talk about the integrated energy system, you have electrons that can travel at the speed of light. You also have this gas flow. You have hydrogen flows. They all travel at a different quantity. So that's how you can model the flows over those different energy sources. And also when you're coupled, when you integrate all those systems together, so you can look at the overall picture, the combined picture. So when you combine different sources, different components, how can you evaluate the overall performance of the combined system? And what's the reliability and impact of those hydrogen on your traditional power system on the other side, from the gas system side, how we can evaluate that reliability as a result of integrating a lot of hydrogen into the overall picture here. And also all those electrolyzers, they have different operating states, which I will talk about later. They have different feasible regions that the region can be operating in. And also they have technical constraints, especially with electrolyzers. So all those needs to be taken into account when you solve this problem. And also when you blend the hydrogen into the system, it's going to create this reliability problem. And how you can optimize the overall system, how you can optimize the planning from the planning from the operation point of view, how we can optimize the entire process so that you can have this robust power gas hydrogen integrated energy system. That's a PGA. That's what we call it. OK, so we have those different topics. So then we are trying to address each of those through this seminar. So the first one is a reliability assessment of integrated energy system considering hydrogen. So this is basically a system configuration that we have here. So we have a green hydrogen is, as I mentioned, is produced by renewable energy with zero CO2 emissions. So we have those green hydrogen CO2, which does not have any CO2 emissions in here. But electrolyzers are going to use the energy from those green energy sources to generate hydrogen. Hydrogen will be put into hydrogen storage here in the middle. Can you see the mask? Yeah, OK. Here. Let me see. It's into energy storage here. And this hydrogen can be used through this machination process to generate this successive natural gas. So this natural gas can be blended into the natural gas system that we have now. I mean, there's no requirement for additional infrastructure extension to use hydrogen for now. You can just blend it into the current pipelines that you have. So as you can see, this is one way using this is a dominant way of using hydrogen now. So I mean, before you do this, so you can see that in the past, you have this natural gas system only. Now you are blending another different gas into the natural gas pipelines. How is that going to affect your overall system performance in terms of reliability? So that's what we care about. The reliability here refers to the level of reliability of the power system overall load. The load can be an electrical load. It can be hydrogen load. It can be gas load. So when you're curtailed those loads in some scenarios, you have a reliability problem. So we want to maintain the highest level of reliability, meaning that we want to maintain no load that is being curtailed at any time as best as we can. So that's our goal here. But you have this overall system, then we're going to evaluate the reliability of this command system. So there are different ways to monitor to model this gas flow. So normally, there are three type of equations. One is so-called state equation here. Basically, state equation is used to establish the relationship between the gas pressure, density, temperature, and the other gas and type-related physical properties. So you can see basically you can measure. You can use the average molecular weight of the gas you have. And then the gas can be a mix of a gas sources. So you have gas fractions here. And it can be a mix of hydrogen and a natural gas. And then also, you have this compressibility factor, meaning that how the gas can perform from the real gas. And also, you have the equations that can represent the temperature pressure. These are all the important characteristics, of course, of these natural gas pipelines. So when you model the gas flows, you have to consider all those factors, temperature, pressure, et cetera. So when you blend the hydrogen into the overall mix, all those parameters are going to be changed. So because you are blending another source of gas, so that's the reason that the state equation has to be changed. And another equation is called a continuity equation. So because it's a flow, it flows continuously. So you have this continuity equation, which basically ensures it's a principle of conservation. If you assume the temperature and the compressibility factor remain constant, which is the general assumption we take in this kind of calculation, you can use this partial differential equation to represent this continuity equation. We also have another equation. It's called a momentum equation. Basically, this represents all forces acting on the gas particles. So here, we assume the pipes are horizontally placed. And also, there's a very limited solution area, a solution with the inertia and the kinetic energy. So if you also use this volume metric gas flow rate, F, then you can simplify this equation as this. Still, those two equations, they are partial differential, randomly solvable. So then you have to figure out the way how to simplify that. So the way we do this is through this back difference So you can use those differential equations to approximate the equation that you saw earlier on the other slide. So once you do that, so you can simplify those equations. Those equations can be much easier solved. And also, you have this continuity equation, which is also simplified. And also, the equation, and also the LAMPAC issue. LAMPAC means that you have, because natural gas, you're going to have some residual natural gas in the pipeline. So because when you take the gas from the pipeline, there are certain amount of gas that are still being stored in the pipeline. So that's called a LAMPAC. You can think of that as a storage capability. So you have certain natural gas being stored in the use. So that's kind of the model gas stored in the pipeline. And it's very important for emergency conditions. Because it can be used as a storage kind of a device. So when you need that, you can use that as a residual gas for those emergency conditions. And so you have those three equations now. And also, if you look at it, if you discretize the gas flow by dropping this continuity equation and the LAMPAC flow equals off-flow, you have this steady state gas flow equation. Basically, this is a simpler version of a dynamic gas flow than I introduced earlier. So this steady state gas flow is a simplification of a dynamic gas flow. So you have this backward difference equation as shown here. And you have momentum equation shows here. So this is a word that's mostly computational. So you can solve those steady gas flow more efficiently than natural gas than dynamic gas flow. But this will introduce errors, which we will talk about later in our results. So then you have three sets of equations, those are state equations, continuity equations, and the momentum equation. And you also have two versions of a gas flow. One is the dynamic, is the steady state gas flow. So that we're going to look at how those are kind of forming modeling techniques will perform your results and how we can simulate the systems through those equations. And all those equations, I mean, some equations can be even further simplified, like a momentum equation can be even further simplified by linearizing those equations. And those steady state gas flow models will provide this benchmark that we're going to compare against. So you have this benchmark, the basic scenario is the steady state gas flow, which is what people normally use now. You don't consider any dynamic. Those continuity equations, you don't consider lambda back. You have the steady state gas flow. But this is the most coarse way of modeling this gas flow. But we're going to compare with this. So then we have this overall problem to minimize overall renewable energy curtailment and overall power and gas curtail. So you can see that we have this minimizes this overall energy in our server and also subject to a number of constraints. We have power to hydrogen and the methane constraints. You can see a different constraint here. We also have a power system constraint tool. We have a natural gas system constraint. So you have a number of constraints coming across different systems in your overall optimization problem. So then you define this. So these are very commonly used in power system industry now, especially in the top two. The first one on the power system side is called the loss of load probability. So basically it tells you the probability of you losing some load. And then you can define that at the system level and the bus level. And also you can have this EDNS expected load supply. We also define at the system level also at the bus level. So this tells you how reliable your system is. It tells you how good your system reliability is. And also you can define those indices similarly for gas system tool. You have a loss of gas probability and the expected gas not supply EDNS. And for the wind farm, you have a loss of a renewable. We're just using wind as one example here. Loss of energy probability. And the amount the expected are not supply ERNS. So we're going to run a number of cases trying to calculate all six different indices. So this is the way we run the simulation. So this is a flowchart of that. So we use sequential multicolor. So basically you run the system. Because we're talking about the reliability, you run the system over a long time horizon. So here, our system over a year. So you have an entire year. And then you cannot run the entire year all at once. You have to kind of divide and conquer. You derive the entire year in our case into 52 weeks. So you run one week at a time. When you're first week, then you'll run a second week. You'll keep going. So you can see that this is the loop that represents the way that we roll the simulation by week. You're kind of the first week, and then you go to the second week. We solve the problem using this two-stage optimal energy shining approach. So in the first stage, we got this harsh solution, which is the steady state of gas flow. The second stage, we're using this small detail of the dynamic of gas flow. So then you run this loop for a year. And you're going to calculate the variance of anything that's served, this data that we have here. So once you have the threshold delta mean, then if the calculation of delta is less important than you'll pre-define, then we say the competition has converged. Then you can generate all of those six availability indices. So this is the test render we are using here. And this is a synthetic system. Basically, we don't have a real system now. And so this is the HB24 bus power system. We also added that 20-node natural gas system into the mix couple of the system now. And on the power system side, we have on the power system side, over 22 conventional generators, three wind farms with P2 hydrogen and methane systems. And we have three wind farms here. And we have natural gas system here. So also we model the outage for each component in the system. So we run every detailed molecular simulation for every component in the system. So when you look at our cases, we want to see that through kind of a detailed case study, how can we see the different reliability performance under different scenarios, under different modeling assumptions? So you can see we have three cases here. First case, C0 is a dynamic gas flow with P2HM and hydrogen effects. Hydrogen effects is where you have those physical characteristics change by your blend of hydrogen into the gas network. So we have a dynamic gas flow with P2HM and hydrogen effects. That's our proposed model. I mean, presumably, it's going to be the best model. And then the second case, C1, is a stainless gas flow with P2HM, but no hydrogen effects. You'll ignore that effects. And then the third case, C2, is a hydrogen gas flow with hydrogen effects, but no P2HM. Here, and you can see that all those three cases can convert using this picture here, over 30 run, 30 iterations. They can all convert. The best reliability performance is observed in the proposed model. So we can see that's the most accurate model. And then the worst L1P, L-O-G-P, and the variances occur in C1. And then the worst L-O-R-P, and the E-R-N-S occur in C1. So they have different performance in different scenarios. So then we want to look at the impact of a gas flow dynamics. So here, as I mentioned, so we really argue that you should use a dynamic gas flow that better capture the flow, the movement of the gas in the pipeline. So I know we compare that with the state natural gas flow. We should not consider the continuity equation and the LAMPAC impact. You can see between the C3 and the C4, you have very dramatic difference in terms of L-O-N-P, in terms of L-O-N-P, very different performance across different reliability indices. So now you can see that the takeaway here is this, if you use a different model, you're gonna have different results. So you should try your best to use the best model, most accurate model, right? So that's our goal here. So here we are saying that through this number of detail, the case studies, it does give you different results if you choose to use a different model here. You can see that in C3 and C4, we have different performance and across different scenarios. And also this is where we increase the hydrogen fraction. So there's a 100% system in the first place. Now you are adding hydrogen into the mix, right? So you don't want to add hydrogen all at once. You want to add hydrogen little by little. So we are adding that by 5% at one time and increase that to 10%. So here this is what we are doing. Again, as you can see, if you use a different representation of gas flows, then the performance of a reliability index will be much different. You can see that this is a different, those are blue curves and in those red curves. Okay? You can see the difference here. So again, the key point is that you should strive to use the most accurate model as possible. That will give you a totally different picture. As you can see here, if you use a C3, which is a dynamic gas flow with hydrogen effects, the IOGP goes up. But if you use a C5, which is a dynamic gas flow without hydrogen effects, it actually goes up. Totally contradictory conclusion. So something you should be mindful of. So given that you know all those reliability effects by using hydrogen in the system, how can you optimally schedule and operate the innovative systems? So here we want to look at this more comprehensive system configuration here. So here you can see that we are adding heat. Because the heat is often the byproduct of the overall process here. As you can see, you can use the hydrogen renewable to generate hydrogen, green hydrogen here. And that's the water system also comes into the picture. You have water here and the water circulation system. And you can heat through here and to provide this heating to the district customers. Then you have this electrolyzer. And the hydrogen can be put into storage and that hydrogen can be directly consumed by hydrogen consumers right here. And also hydrogen can be used through this machination process to be blended into natural gas system. You can see hydrogen is kind of a coupling factor here. It connects with water. It connects with hydrogen customer. It connects with renewable. It also connects with the next system here. So this is the system that we are looking at. We are trying to optimize the operation of such integrated system here. Okay, so then we throw in a bunch of formulations here. So we have this combination factor here which basically acts represent the on and off status of electrolyzer. So if it's on and this one, if it's off, it's zero. Then you can see that we're trying to the feasible region of electrolyzer using this convex methods. And then we have this combination factor. We have the power of balance for electrolyzer, powerful hydrogen, powerful heat. Also, we have different constants of such as electrolyzer performance. This is for the temperature. This is for the unit commitment on and off the equivalent for those electrolyzers. That was the hydrogen integration constant. So we have powerful hydrogen, hydrogen storage constant, hydrogen consumption balance constant. Also have this conversion between hydrogen and to methane as you can see here. We also model the conversion efficiency through this equation here. And also hydrogen can be blended with methane into the system too. As you can see, this is the amount of hydrogen into the pipeline. And this is the methane that you already have. And this is the total amount of gas that you have as a result of this combination. And also you can represent this heat transfer process here which is a capture of the change of electrolyzer temperature. Also you can capture this heat loss through this equation here. So we have those that detail if we're going to capture this electricity, also capture usage, utilization, and appearance constants, also we can capture the heat transfer process here. So the overall objective function is to minimize the overall cost. So this cost here can include the UC, the unit commitment cost on the traditional power generation side because for every power generator, either it's a gas generator, coal, whatever you have. And associated with when you start those generators from the beginning, it has a stopper cost. When you start them down, they're shut down costs. When you operate them, there's a breaking cost associated with those. And also you have this natural gas cost too. So because when you burn gas, it incurred the cost. And also we have this load not served scenario. So when you have this load not served scenario, you worry about the worst case. So that's the reason we use the kind of a robust optimization representation. And here we are saying that what's the worst case that you'll lose most of the load? And what's the load can be gas, can be heat, and can be power. So when you calculate all those, so your overall objective is really to minimize your cost under those worst case scenario. So we want to try to reduce the power being shed in those worst case scenarios. And also also we want to reduce overall operating cost. Again, we have a number of constraints here. There is a power to hydrogen and the heat constraints. And also we have a natural gas network constraint. We have a distribution network constraint which maintains the nodal power balance. We use this this flow model, node voltage and the UC and linear design rules and then for the heat side, we have this nodal heat balance constraint temperature of the supply types and the temperature at the conference nodes. Okay, so when we run this HB33 bus case, which is a slightly bigger than the previous case. And here you can see that we are adding this heat network into the picture tool. So you can see those are red curves, red dots. Those are the heat pipelines and heat networks, heat demands. So you have this natural gas pipeline here, which is a representing blue. You have hydrogen, you have a power in black and then you have a heat here. And also we have renewables into the picture here. So if your screen is hard enough, you can see this wind farms right here. Here, and there's a wind farm right here. Okay, so now we have this integrated system that has different energy flowing throughout. Okay, so again, we run a large number of cases. Basically, that how each of the case perform and which one is the best kind of configuration that we are seeking. And for the S1 to S2, we have a steady state gas flow without hydrogen effects similar to our previous case study of SH1 to SH3, which is steady state gas flow with hydrogen effects. T12, T3 and a dynamic gas flow without hydrogen effects. And TH1 to TH3, dynamic gas flow with hydrogen effects. So for each group, and we slightly increase the hydrogen fraction from zero to 10 to 20. So we're trying to increase the fraction, the penetration of hydrogen into the gas network. You can see that the total overall operating costs decrease with the increase of hydrogen because when you add a hydrogen, you'll reduce the amount of gas you are burning and has high costs. So when you add a hydrogen here, and especially hydrogen is produced by wind which is zero operating costs. So hydrogen comes at three and no cost to you. So when you increase the amount of hydrogen in the overall gas pipelines, you are burning more hydrogen and the less gas you are reducing the overall cost, which is the benefits of using hydrogen here. And also the steady state of gas flow leads to different results. If you look at this table on this side, they generate a different generation dispatch results. Those are the generators, command cycle units, electrolyser and gas well. And because you use kind of a coarse resolution gas flow model, you're gonna have different results from the more accurate dynamic gas flow model. And also if you neglect those hydrogen effects, meaning that you ignore that's a physical property by blending those hydrogen because as I showed you earlier, so that's gonna change a lot of parameters that are used to describe the gas flows. If you blend hydrogen in there into the overall mix, you don't change those properties, those parameters. Now your results are going to be misrepresented. So that's another point. So then we have, look at this reliability side, operational side, one kind of planning side. I really think, myself, really think that this kind of integrated might be our future. So you are not going to deal with hydrogen system alone and you are not going to deal with gas system alone. So in the future, it's gonna be combined as a system. And especially when you have hydrogen in the mix, because hydrogen could be a game changer, could be this coupling bridge between different energy sources as I mentioned earlier. So when you have this overall integrated system, that's what we call it, kind of hydrogen half. So all the other system can be centered around hydrogen, because hydrogen can be used to convert a two-way back on the frost between different energy sources, between electricity, between gas as I showed you earlier. So hydrogen can be the middle man here. So that's the reason we call it hydrogen half because it's very convenient to be used because hydrogen can be easily blended in the gas network, can be transported through trucks and can be generated locally. So if you distribute your electrolyzer across your territory, then you can generate the hydrogen locally. It's like a DR, right? So you don't have to generate hydrogen in a centralized plant, which is not a benefit for you being local. Also, it can be transported too. And also in the future, we can have this high-temperature electrolyzer, which has a much higher efficiency than the current low-temperature electrolyzers now. The current efficiency is 40 to 60. In the future, if it was high-temperature electrolyzer, the efficiency can be much higher. And also, it can be connected to the external power system here too. So here, we are specifically looking at the solar power plants. You have these CSPP plants here. CSPP plants, we have thermal storage because it's going to generate heat. You have the solar field that has taken solar energy as input to generate the heat and generate the power. Heat can be put into a thermal storage here that can be used to supply your heat demand in your district. Also, it can be used to generate power through this power block, okay? That power block can be used to generate the power. Electricity can be supplied to power system. And those high-temperature electrolyzer can be taking power from the external power system to generate hydrogen too. So you can see that they are kind of an intertwined system here. And we have also another set of different equations to formulate all those constructs, associated with each of the sources. So we have a CSPP constraint because you have thermal here, different from those PV panels, which does not deal with heat. We have CSPP here that also has a seat storage and generating capability. So we have seat thermal power balance constraint as shown here. And we have a power block output constraint. That can be used to represent the power generation processing here. We have thermal storage here. This is the state of that storage device. And then we have a range of storage, upper bound and lower bound. And then we have these high-temperature electrolyzers here. So those high-temperature electrolyzers will require thermal power input. We will require electric power input. So those are represented through those two equations here. And also they are going to generate the power. And for overall operation of a high-temperature electrolyzer, you have power for hydrogen, and you have reserves, and you have stop and standby constraint, and multi-state UUC problem. Here, the key point here is that so for those high-temperature electrolyzers, they have another state which is not on or off. It can be run on this standby mode. So it can be sitting idle there. So we have this multi-state UUC unit commitment problem. You have a zero, one somewhere between zero and one. So the standby, you can hear it. And there it can be all represented through those equations here. And then, as I mentioned, so this tells you this kind of high-temperature model that I was talking about. So only, this is your electrical bus. So you have external power grid connecting to this bus. And that can be connected to this bus too. Also you have an electrical demand also connected with this bus. This is your electrical bus. Also you have this hydrogen hub bus which is sitting in the middle. And then this bus, this hub, it can be connected to hydrogen network, electrolyzer. And then electrolyzer is a bridge between the electrical bus and hydrogen hub. Also you have fuel cell. Fuel cell. And also you have hydrogen demand here. And on the lower side, you have this natural gas system. We have natural gas network, natural gas demand. As I mentioned earlier, natural gas can be used to generate hydrogen through this SMR, steam-methane reforming process here. And also hydrogen can be used to generate the synthetic natural gas through machination. Okay, you can see this overall picture here. So for the gas system, we use those equations here. And we can see those are the top two. They are used to represent this steam-methane reformer. How to use the gas to generate hydrogen. Also you have hydrogen storage equations. These are the ranges and these are state-of-the-charge. And then we have those low temperature electrolyzers too. So then we have hydrogen pipe, upper bound the lower bound, hydrogen demand, hydrogen to power to hydrogen from hydrogen equation and reserves. And those are the hydrogen balance constant here. So those hydrogen balance constant, because hydrogen is like electricity. For electricity, you have a Kirchhoff voltage law. You have a Kirchhoff current law, right? So, and the hydrogen is the same thing. So you need a conservation through up. So you have those hydrogen balance constant that represents the principle of conservation here. You have exchange of the power, you have exchange of gas. Between the hub and the external system. Okay, so because we are talking about this investment, planning problem, so we are talking about investment cost. So all those detailed constants are going to this formation here. So all those detailed constants are going to minimize the overall investment cost. So investment cost includes this generator and the line investment cost. When you put up a generator, it has this capital cost. And then you have this SMR and a few cell investment cost, initializer investment cost and hydrogen storage and pipeline investment cost. You also have this operating cost. So when you do your planning problem, you don't only focus on this investment cost. So you have front investment cost. Then when you're looking into the future, you'll need to simulate the operation under each investment scenario for the future years, right? So this is where you calculate those operation costs. So those operation costs include the generator operation cost, power shining cost, hydrogen operation cost and hydrogen shining cost. And also guess what? And also spinning reserve cost. The system planner and system operator, you always maintain certain extra redundant capability in the system, in anticipation of something that you have not foreseen in the first place. So those are the reserve are useful. Reserve can be used for power system, can be used for gas, can be used for heat, can be used for any other energy sources here. So these are common measure that we take to what is these scenarios? Those are the reserve costs. Because you are accommodating those reserves, it's going to incur this spinning reserve cost. So you need to consider all those costs, when you do your investment planning problem. So then we consider this a problem, we have power system, hydrogen, and all those equations, then every candidate has different locations, sizing and the energy issue, right? So when you do your planning, either for power, for hydrogen, for gas, you have different choices. You have this SMR, you have electrolyzer, you have storage, you have fuel cell, your pipes, you have generator, you have a transmission lines. For every components, you have the challenge of picking the location for that particular component, picking the size for that to pick the timing when you build that. If your planning horizon is 10 years, you build it in the first year or in the second year or in the last year, right? Also the types of technology you're going to choose. So all those have to be considered in your investment problem too. So those are the investment costs that you have here. So you can see this overall objective function is to minimize the overall cost. You're going to minimize the overall investment cost creating cost. So the planning constant in location, size, and type, you're going to pick from a certain budget that you have. And also you have those operation costs. All of the details of the constant that I showed you earlier, CSPP and hydrogen hub constant, power system constant, and then network gas system constant. Okay. So then we run this integrated system planning using 24 bus case and 20 node natural gas system and four hydrogen hubs. And you can see that the four hydrogen hubs are sitting in the middle here because they can be used at the bridge between different energy systems. And you have a power system, hydrogen hub constant, you have two types of SMRs, I have electrolyzers, fuel cell and hydrogen storage. I have a natural gas system, I have different two number, two gas well, I think pipes and the two compressors. And we have demand because of power, gas and the hydrogen here. So then you have a demand that is known to you. So then you need to plan a system to meet that demand. So now we have three cases here. A1 is the proposed a specific programming based coordinate planning model. Here, we look at different scenarios because the future demand fluctuating can be forecasting. So we use the storage, we use the specific planning methods by assigning a probability to each individual scenario. Now you look at the overall picture trying to get the lowest expected cost. The second approach A2 is only low temperature, there's a low efficiency, low temperature electrolyzers are used. And A3 is uncalled in the planning model without electrolyzers themselves. So then those are, this table shows you a different performance in terms of cost, in terms of choice of technology. As you can see here, A1 has the least total cost because that's the most optimized way of building your integrated system. A2, A2 is gonna increase the overall cost by 43 million. And A3 has highest total cost because it totally uncorrelated or totally not totally optimized. And also you can see that 38% of wind is gonna be curtailed in case three. So that's very bad scenario, because it's not coordinated, it's not co-optimized. And those are the key points here. One of the key points is that you can see that the electrolyzer can really help the system because they can help or absorb the renewable energy. When you have surplus wind and surplus solar, which you can not use, you can use that surplus wind and solar to generate the start for long duration of time. Because hydrogen, another benefit of using hydrogen is that it can be started for a long time. It's not like a battery. If you start energy, the battery is gonna die out, right? So for the hydrogen to put in the container in time, it's gonna be there for a long time, which is good. And so those electrolyzers can really help you. Also, those are fuel cells that can also offer additional electricity of hydrogen. You can use hydrogen through fuel cells to generate the power. Also, those are fuel cells, they are inverter-based resources that can also provide great support to the overall system. So, and then we went to kind of look at the different modeling techniques associated with those electrolyzers. As I mentioned, so we have the different modeling kind of formulations. You have one state, just two states, B two states, just on and off. Three states on and off and stand-back states. Also, you have two types of electrolyzers. One is the low-temperature, low-efficiency. The other one is that you have both. So as you can see, we run the case for different numbers through here that the high-temperature electrolyzers can reduce overall cost because it has a higher efficiency of running. Our proposed model, those small detail operating a state model of those electrolyzers can reduce the total cost and it's more consistent with overall practice here. And we see the improvement over the current state of light. And so we're trying to see that if our massive can scale to a larger system, then we go for this 72 bus HV case. Now we have a 40 node natural gas system in half. So it's almost double the size, more than double the size of the previous system. So we run the case. Again, we run different number of cases and we got a similar conclusion here. And you can see that C one, which is our proposed measure, because you have the least total cost and C two and C three will increase the overall cost by 75 billion and 436 billion. And then the curtail wind is still very high about 25% right here. And also a model can take a longer time to run because there's a lot much bigger system here. It takes about 20 hours to run, but we should still acceptable because you are running, there's a planning problem. It's gonna take a year to build. It's not like an operation problem. You'll need to run it in real time. You'll need to get the results in the next five minutes. But for the planning problem, it takes years to build. So this and yeah, so in conclusion that I showed you that the work that we have done recently, we are learning this too because hydrogen is a new topic to us, but not many people have looked into how we can integrate hydrogen into overall energy system operation. And then I think we are the one of the first to look at some reliability issues associated with such an integrated energy system, how you can run your optimal operation plan your system in an optimal way for this such an integrated system. I hope you have learned something from here. And so going into the future, that we have a lot more work to do. And you can see that we can do, we can enhance this reliability assessment. And we can do our optimal operation better. And also we can do this optimal planning better. So there's a lot of things that we want to add to this tool and we want to have this model bus optimization model that can consider this uncertainty from wind and solar better. Because if your wind and solar is a source for your hydrogen production, the solution wind and solar is gonna affect your overall hydrogen production in a big way. So we need not only that in much detail, that can be added into the future. And also we did not consider any transportation. As I mentioned, so hydrogen can be put into these trucks, right? That can move around, that can move hydrogen from one point to another point. That has to involve this transmission networks, modeling constraint. And we don't consider any of that here. And that's gonna throw in another dimension of difficulty and complexity into that overall framework that I'm showing here. Yeah, thank you. I hope you have a good something from here. Thank you.