 Hello everybody. Welcome to the New Mexico EPSCORE Smart Grid Center webinar series. We'll get started precisely at noon. Thank you for spending your lunch hour with us. My name is Ann Jekyll. I'm the Associate Director of New Mexico EPSCORE and on behalf of myself and Bill Mitchner who is the State Director and Principal Investigator of the New Mexico Smart Grid Center. I'd like to welcome you to this series. This is the first webinar that we will continue throughout the academic year. The webinars will be the fourth Friday of every month at noon. So please mark your calendars. And the purpose is to present research and promote collaboration across our project. But we will also provide a venue for outside presenters working on subjects related to microgrids and other topics of interest to the New Mexico Smart Grid Center to present. If you are interested in presenting or you have a suggestion of someone who should present, please contact me. You can find my contact on the website pretty easily. An item of housekeeping. If you would like to provide any comments or questions, there's a question and answer box in the Zoom webinar interface. And that's the best way to interact. So if you have questions throughout the presentations, please enter them there and we will address them as time permits. Our next webinar is lined out. It is October 25th at noon. Gary Oppendahl from Ameri Technologies will be presenting on networked DC microgrids. Ameri Technologies has a couple of microgrid demonstration projects ongoing in New Mexico. And this will be an excellent opportunity to hear about them and find out opportunities for collaboration and what some of our industry partners are doing. And while you are all here, I'd like to just let you know about some upcoming dates. I won't go through all of these, but I did want to let you know that the first two here, the Jack Davis and Sandia National Labs grid resilience forum, you can attend remotely. So no matter where you are in the state, you can attend these events. And also we encourage all of our undergraduate and graduate students to submit an abstract to present either an oral or a poster presentation at the New Mexico Academy of Science Research Symposium. Those abstracts are due October 6th. More information about everything is on our website, which is nmepscore.org. You'll see here the news events and webinar links. So without further ado, I will kick off year two of our project and this first webinar with an overview of our center. We have many new students who are joining the project right now and others are just starting their scope of work for year two. So we wanted to reorient you and contextualize the project. The New Mexico Smart Grid Center is a five-year project funded through the National Science Foundation EPSCORE program that establishes an interdisciplinary research center to address the design operational data and security challenges of next generation electric power management. And specifically our project is looking at how we can transform existing electricity distribution feeders and turn them into a set of interconnected distribution feeder microgrids. Some of you are familiar with the power sector and some of you aren't, so I'll take a minute to define the distribution feeder. That's the final stage of the electricity distribution system. So in a traditional utility grid, we've had large centralized power plants. They generate power. It goes through high-voltage transmission lines. It's stepped down at distribution stations and then delivered to customers via the distribution feeder. A microgrid is an interconnected set of load and distributed energy resources that can operate independently of the grid. So at the microgrid controller, you can island it from the grid. And this is a discrete entity also. So we can take the idea of the microgrid and turn the distribution feeder into a microgrid. Why would we want to do this? Well, the answer to that is contained in our project's acronym, Sustainable, Modular, Adaptive, Resilient, and Transactive. I'll take a minute just to dwell on the resilient and the sustainability aspects. Resiliency because the microgrid can operate regardless of if there are issues with the greater transmission grid. You think about Puerto Rico after Hurricane Maria, the areas where there were microgrids were up and running much more quickly than the greater power grid. Some people were without power for over a year. So that resiliency aspect can be very important, whether the grid issues are human-caused or a natural cause. And the other is the sustainability component. As we move toward a lower carbon future, microgrids can allow for more renewable energy and energy efficiency technologies to be integrated into the grid. And check out how happy these people are. I think that if we were on distribution feeder microgrids, we would all be that happy. We have four core research areas that are part of our project. And I won't dwell on them here because you'll hear about them in more detail from all of our presenters. But very quickly, the architecture group or research group one is looking to optimize distribution feeder microgrid design incorporating human preferences. The networking group is creating networking and communication systems that are scalable, secure, and protect user privacy. The decision support crew is integrating machine learning, data mining, and knowledge-based techniques to make computer-aided and data-driven decisions in the smart grid. And deployment is where we tie this all together and take models, simulations, and deploy them in diverse testbeds across New Mexico that you'll hear about today. This project is bounded by cyber infrastructure and human infrastructure. We've started referring to this in the office as the EPSCORE hug. But we have a high-performance computing resources at New Mexico State University that all of you can access and also some data archiving and sharing resources run through UNM. And the human infrastructure refers to training students, hiring new faculty, and all of the training programs that we have as part of our project. We are focusing on research here today, but I would be remiss if I didn't mention our education outreach and workforce development programs, of which there are many you can learn about all of them on the website. Everything from training two- and four-year college students and research experiences to science communication fellowships. There are many opportunities for you to engage. Make sure to read our newsletter each month and you'll find out how you can sign up. And for those of you students that are just starting at your own institution, I wanted to make sure that you understand the scope of this project. New Mexico EPSCORE and the New Mexico Smart Grid Center bring together all three research universities. The two national laboratories, community colleges, Explora is our primary outreach partner. And there are opportunities if you want internships at different places. If you're an undergraduate student and you're looking to get a graduate degree faculty that we can help connect you with. So keep that in mind. With that, I will turn it over to our first presenter. Olga Lavrova is an associate professor in electrical and computer engineering at New Mexico State University. And I'll also just remind you if you have questions, as Olga is talking, to enter them into the question and answer box. Olga, take it away. Okay, thank you and for the introduction and for the overview. So I will present some of the research work going on as part of research goal number one, which develops architecture. And this work is a work of many team members. We primarily consist of faculty and students at NMSU as well as faculty and students at UNM. Some of the students have graduated. Some of the newer students are joining this semester and this year. So I will not go through the whole list. But these are all good opportunities for any new students to start getting involved. We do have industrial partners that are listed at the bottom. We see them as very important partners in this whole project and you will start seeing more and more of their participation as we progress. So next slide please. So just as Anne introduced in the introduction, our goal is to start looking at distribution feeders as a new component for our electric grid infrastructure. And there is a strong motivation for us to start thinking about this as a completely different component. And the motivation being that we are progressing into very significantly different mix of energy generation resources, whether renewables being at the forefront of this generation mix. Additionally, we are all adopting electric vehicles and other electrification of other industrial sectors. The picture that Anne showed at the beginning showed an electric vehicle. And those are beginning to be more and more widely adopted by us as consumers. So we really need to start thinking about what other required changes that our distribution feeders may need to start facing. And so I have several key questions listed here on the left. As you read through those questions, I encourage you to think about some of those questions. For example, where should we deploy all of our renewable energy sources? Should they be deployed randomly wherever each individual home owner decides to purchase a PV system or an electric vehicle? Or should there be some bigger picture and bigger plan that should enforce those? How do we protect safety, security and reliability of our new electric distribution feeders? And who is responsible for that? Is that still utilities responsibility? Is that now individual home owners responsibility? Is there a mix of those? And how basically do we define that as a proper architecture? How do we make sure that the grid still stays safe in terms of protection? There is a lot of work that goes on in the electrical grid that is not visible to us as traditional consumers. For example, protection by fuses or protection by relays and circuit breakers is typically not seen by traditional consumer. But this is something that needs to start changing as we adopt more and more electric vehicles and renewable energy sources. So with this, I will go over some of our current achievements and plans. So next slide, please. So as one of our accomplishments for the first year, we have graduated several students and I will use Gazeta Varys' dissertation as an example of one of our really interesting achievements. His dissertation centered on decentralized robust optimization for transactive scheduling of distributed resources. This is obviously as applicable to our distribution figures. And the question of his thesis was, how do we take uncertain solar resource shown on the upper right as a series of graphs of solar irradiation throughout the month? How do we take this typical uncertain and somewhat unpredictable solar irradiation? And how do we optimize all of the energy transactions that may happen on one of the distribution figures? So picture on the lower left shows a six-post test system. This is how power systems are modeled in power systems transmission and distribution infrastructure. Each of the circles represent a traditional or a green renewable energy source. And pictures on the right, middle and bottom, show mathematical formulation of how each of this optimization of energy transactions can be formulated. I understand the numbers are very small, but you can see that we are representing some of the loads such as electric vehicle and some of the exchange of the electrical energy with neighbors on the same feeder as those arrows that go between customer 1, 2 and 3 and basically provide bi-directional energy and information communication. So this was a very interesting and very useful formulation of this mathematical notation that can now be extended to much larger systems. Obviously, there are many more than six customers on each of the feeder. So we are looking at how to extend this mathematical notation into hundreds and thousands of customers. Next slide please. Another example of our accomplishments is looking at how do we actually interface with real hardware on the grid. And by real hardware, I mean PV system with a traditional inverter or any other of the components that actually do produce electrical work and electrical energy conversion. So the picture on the upper right shows a similar notation of a certain bus IEEE electrical standard system. And we are looking at things like how will this hardware respond in case there is a fault or a problem as indicated by this yellow short circuit symbol. And so we are modeling how some of the traditional and new advanced smart inverters will respond to these issues and how will they protect the consumer and how they might isolate and island the whole system. Next slide please. As additional accomplishments we are working on, we have already published several publications and we are working on many more. Additional interesting aspect of our work is that we are also looking at how these distribution systems will work with transmission planning. So paper on the right shows our exploration of transmission planning for different objectives and some of those objectives include being able to function as an individual microgrid or islanded microgrid. Next slide please. As part of our work we are also looking at some of the other hardware deployments and as I mentioned we are working on some of the hardware in the loop interfaces. Figures on the right show some of our lab installations of hardware in the loop inverters, microgrid cabinet and our PV system plus storage system on the roof of our own electrical engineering building where we are collecting the data from this PV system and we are implementing algorithms that I described briefly from the thesis by Hazel Abaris. Next slide. On UNM side we also have strong contributions from the UNM representatives that I mentioned at the beginning. They did a lot of good work on looking at every use cases that will form some of our distribution feeder microgrid use cases going into year two. Additionally there was a survey that was administered looking at how traditional customers, traditional electric utility customers might respond to some of those changes. For example, will they be interested in adopting those changes or will there be some resistance? What kind of incentives or what kind of dis incentives should be employed so that customers are interested in participating in this infrastructure? That survey has already been administered in June. Additional survey is in the planning stages right now to be deployed a little later during year two. And also there are additional publications in terms of journal papers that have already been published or still under review. Next slide. And so I keep on alluring that there is a lot of work planned for year two activities. As you read through this slide I'll just mention some of the highlights that we really need to start building up on the year one work where we have done a lot of characterization of different variable parameters. We built mathematical models and now we need to actually start putting those models to work. Additionally we will start looking at how do we define metrics for resilient microgrid and what does that mean. Also what does a sustainable microgrid mean. And then we will start looking at how do we combine those mathematical models resilience models resilience metrics and basically look at different optimization of different microgrids. And with this I will hand this all back to Anne. Great thank you Olga. And we had a question that came in while you were talking and it was could one of the professors and I think that will be you please explain how a distribution feeder microgrid is different from a typical microgrid. Sure should we address the questions now or at the end. Why don't we just address that one right now. Okay sure. So there is a slight difference between distribution feeder microgrid and traditional microgrid. And when I said slight differences that could be the size whether it's in terms of energy and load balance. Distribution microgrid may be smaller than traditional microgrid by the virtue of the fact that we are only looking at the residential customers not heavy industrial customers. However some of those boundaries are really somewhat blurred and what we're really looking at here are distribution microgrids for distribution infrastructure that is already in place and some of the future developments. But we are not looking at larger microgrids like industrial parks or commercial customers. Great thank you. Okay I would like to introduce our second speaker is David Mitchell from New Mexico State University. He's an assistant professor in electrical and computer engineering and also want to recognize him for just being awarded the New Mexico EPSCOR excellent mentor award. So congratulations David and please go ahead. Hi good afternoon everyone thank you Anne for the introduction. Yeah okay so the goal of this research is to design a comprehensive network architecture that will complement the DFM architecture we just heard about from objective one. And this is generally a difficult problem since existing internet protocol or IP solutions will not scale to handle the data volume and demands of smart DFM. Actually the paradigm that we're going to build this infrastructure on is called information centric networking or ICN. And for those of you that might not be familiar with ICN it moves away from traditional end-to-end type IP network communication and instead of naming devices ICN named data or information. This will actually be a first of its kind DFM network architecture that promises several benefits such as improved efficiency, better scalability with respect to information, bandwidth demand and better robustness and challenging communication scenarios. It's design forms activity one and this will be led by Jay Mishra at NMSU and Michael DeVette Sikiotis at UNM. In order to realize this DFM networking architecture we're going to require three critical supporting activities. Activity two will investigate new and incorporate current advances in wireless coding and compression technologies and this activity is led by myself and Hongkwan at NMSU. Activity three will propose methods to enhance the security level of the ICN based architecture which actually by itself might actually present new vulnerabilities and this will be led by Jun Zhang at NM Tech and Jay Mishra at NMSU. And finally activity four includes the development of novel techniques to preserve user privacy in this data driven network and this will be led by Dong Wang Shin at NM Tech. So in the next couple of slides I'll just give a brief overview of each activity and highlight our your two goals. Okay so I've already talked a little bit about the ICN networking paradigm forming activity one and some of the research highlights from year one include exciting proof of concept results that show named data networking or NDN enables much better convergence and oscillation damping for packet delivery than conventional IP does in DFM networks and this is illustrated in the upper right figure for those of you that know the area. Another area of significant progress was our investigation of the two-way communication impact between electric vehicles and the microgrid and that model shown there in the in the lower right. And building on these results our first goal for year two is to develop a token bucket and prioritize scheduling scheme to be deployed at the routers in the network and this will actually be a critical component to check that the data transmissions conform to say defined limits on the bandwidth and burstiness. Our other goals aim to move our promising theoretical results to our practice including both large-scale software and also hardware in the loop opal RT simulations and these can be used for example to assess the impact and performance of say critical and reliability messages. Next slide please. So we know that the volume of network traffic in the smart DFM may in fact be orders of magnitude higher than the current mobile broadband with stringent low latency and interoperability requirements. In activity two and thus far we've developed sparse graph codes for ultra reliable low latency data communication he's actually akin to and recent developments in 5g cellular and these schemes show remarkable performance when compared to say state of the art color codes and turbo codes shown in the upper right figure. We've also investigated concatenated graphical models for joint data compression and channel coding. Our year two goals in this activity will focus on the challenging extension of these novel point-to-point protocols to and extend those to say the multi-terminal DFM network architecture as well as validating the theoretical models with real and say PMU data. Next slide please. Okay so along with those benefits I discussed earlier the addition of ICN based network architecture to the DFM infrastructure will greatly increase the system complexity and in fact introduce new vulnerabilities entry points and paths that could be exploited by potential adversaries. In year one we investigated efficient authentication schemes for DFM networks including a biometrics enabled multi-factor authentication for electric vehicles and the situation awareness based scheme for smart grid enabled home area networks and that's kind of illustrated in the upper right there. Another research highlight was a novel intrusion detection scheme for DFM shown in the lower right. Goals for year two include an investigation of a so-called identity based cryptography for authentication and methods of achieving secure group communication in ICN based DFM networks that's particularly challenging for us. Then we'll also build on our exciting preliminary results for an intrusion detection system that utilizes semi-supervised outlaw detection and deep feature extraction for detecting cyber attacks using PMU data. Okay next slide please. So activity four I'm here we're talking about privacy and we know privacy will be a critical issue for wider adoption of DFMs and we of course know that fine-grained energy usage can improve the efficiency cost effectiveness and adaptability of DFM operations but it may also cause invasion of user privacy. For example data collected from DFM systems could be used to make inferences about a user's private life such as home occupancy. And your highlight your one highlights in this activity include a Dempster Schaefer based approach to develop a risk based privacy mechanism as well as a comprehensive DFM privacy ontology that provides a standard taxonomy identifying practical privacy factors for each data type and this is illustrated briefly on the right here. Our goals for year two build on this privacy analysis and ontology aim to characterize the uncertainty of these privacy threats and provide a cost benefit analysis for maintaining a given desired level of privacy. Okay and I'll just wrap up here by highlighting some areas for cross-group coordination and the DFM architecture from objective one that's our objective architecture clearly will drive the design of our ICN based networking architecture and so we'll need to study the interplay between the power and information networks exploring joint optimization and control models. We're also proposing to co-model power and information in other prosumer by that I mean consumer and our producer settings especially those involving electrical vehicles in either fixed charging like a charging station or continuous charging such as inductive charging along highways as well as being in different urban and rural settings. Ares of coordination involving objective three that's our decision support include primarily our data driven challenges such as compression and communication of large datasets like PMU smart meter records the feature extraction how we handle that as well as those machine learning approaches to intrusion detection I discussed in activity three. Finally we anticipate heavy coordination with objective four and that's our deployment as we begin transition of our theory and code based toward our software and hardware in the loop simulations and with that I'll wrap up thanks for your attention. Many thanks David. We'll transition then to research goal three decision support and our presenter today is Manel Martinez Ramon who is a professor in electrical and computer engineering in fact all of our professors today are in electrical and computer engineering though our project spans many disciplines including computer science, mechanical engineering, economics to name a few. Manel is the king Felipe endowed chair and professor at UNM and go ahead Manel. Thank you so I'm going to present the research goal number three can you please go to the next slide this is the overview of our research goals and objectives while we are in this research goal we're dealing with data and in such environment we have tons of data this data is non-structure is heterogeneous and it might have problems like lost data tamper data and all sorts of anomalies in the data and in this scenario we have to provide tools to extract knowledge from this data that provides machine learning intelligence to the decision-making tasks in the grid. So our objectives are mainly three the first one is the one in which we attempt to clean and preprocess the data in order to preserve data integrity for example here we want to structure the data we want to detect when the data is missing for example or the data is incorrect and in this in this objective the participants are at Yosia Strada and I'm myself from the UNM and also Huy Pinkau from NMSU and the second objective the second objective is the one which is related to actually provide knowledge to the grid and here we apply techniques to produce anomaly detection these anomalies they can be simple simply data missing or corrupted by malfunctioning systems it can be anomalies caused from data tampering or infractions in the network and we also intend to provide a forecast of the energy and the energy usage and production and also database querying database querying is is very important in this scenario precisely because the data that we have is not structured and it's massive and finally so in this in this part of the project so the participants are myself at Yosia Strada Abdullah Muin from the UNM and Alora Boucheron from NMSU in the third part we attempt to produce this decision making tools and the main idea here is to produce autonomous agents for complex systems that are able to autonomously produce planning technologies next slide please one of the what one of our achievements is in multi-characteria queries as I said before we have we have potentially a lot of data and we want to use this data to answer questions for example where do I get my energy or what is the best service to fix a given problem and this question has been solved and published in a in a conference paper so the next please next slide the the other another achievement or an upset of achievement achievements is related to the use of variation approaches for decision support and we apply this theoretical techniques to range of things for example we try to answer the question of what is the best set of data to train a machine that is able to produce a given forecast and we do it using a probabilistic a probabilistic modeling of of the data we also have different approaches to actually produce forecasts and these approaches they doesn't only produce a reasonable forecast of the energy that we use or the energy that we are going to produce in the next hours or days but we also provide with confidence interval to determine whether our prediction is good or is bad we also are able to study the sky and we do it in order to know whether the clouds are going to let us produce energy or not in the in the next seconds or minutes which we also have probabilistic models of aggregated residential and energy in order to train our models and in order to explain the behavior of the grid next slide please another approach for prediction of solar power is not done by using patient approaches but using deep learning and in this example we have used a deep learning a deep learning machine which is called long short-term memory network and this network it basically observes the temperature due point and other methodological variables and only with this and with provided we have enough enough data for training we are able to do a prediction a very short-term prediction of the solar power generation with a very low prediction error and the results of this research are in a manuscript under preparation by we next slide please another example of our work which is conducted mainly by professor Tilsa Estrada is related to distributed stream analysis for morphing graphs and the idea here is that the IOT appliances can be modeled in order to determine whether these appliances they are working autonomously or they are working in a coordinated way when they are working in a coordinated way sometimes is because of an attack a coordinated attack and then using probabilistic analysis of probabilistic modeling is it possible to detect or predict this this coordination this coordinated behavior and prevent attacks getting this slide please this is another of the examples of our research and this is an application which is dedicated to the detection of anomalies in the network there are many approaches for anomaly detection in in the network some of them they are supervised some of them they are unsupervised and in in this case the researchers they intend to use the very well known super vector machine to do anomaly detection but in a in an environment in which there is a huge quantity of data usually the super vector machines they have an outstanding performance in anomaly detection over many other approaches that are probabilistic but their problem is that they can only deal with limited limited samples and here we have an environment in which this is not true so the researchers there they have introduced a machine that is able to combine the use of a massive quantity of data with the performance of the super vector machine with successful results next slide please and finally we're also working in autonomous agents for complex systems so we have the smart grid is is itself a complex system in which many decisions they have to be taken at the same time and and they have to be autonomous because of a variety of reasons first not always there's not always humans available to take the decisions they don't have all the information and sometimes communications they don't allow to coordinate the the the whole grid to take these decisions and also the complexity doesn't scale properly so the our team is researching in methodologies that are autonomous and they use game theoretical approaches to do this task this night please this is our second year work plan and I have to say that all activities they have been progress as scheduled we don't have any delays and also I have to say that some of the activities that were programmed for the future we already started them we have published eight or more papers and we have some more that are already submitted of 13 committed for the whole project also 10 presentations to conferences and we have several PhDs in in progress also we have submitted several grants related to this to this project then so in objectives one two and three just in order to be to go finishing I have to say that we are going to proceed as scheduled initially without any change next slide so finally for our areas for cross research coordination we are already working with research group one in modeling simulation of faults and tests of multi-objective fall detection systems we are planning to work with group four in data sharing in messa del sol and in prosperity I think it's going to be a great thing is if we have a availability of the data that it's produced in these two sites and I also take the word of research group two that mentioned that we can collaborate in in data processing and and supervisor supervisor semi-supervised learning we are very interesting in that so we will be happy to invite David to our next group that's it thank you thanks manel and thanks to those of you who have asked some questions I see a few questions in the q&a we're going to hold them until the end of our final presenter but if any other questions that you can think of that come up please do put them in the q&a and we will have time to address them at the end so our final presenter is Ali Bedram he is an assistant professor in electrical and computer engineering at the University of New Mexico and he is also our first epscore hire as part of this project so please go ahead Ali thank you and that's my pleasure thanks for the introduction so please go to the next slide so I'm going to talk about the research group four which is the deployment research group before I start about the activities that we have and the plans for the next year I'm just going to quickly talk about this research goal that what is their objective what is the main goal of this research activities that we have on this under this research goal so as Anne mentioned the overarching goal of the distribution feeder microgrid and this project is to address the resilience and sustainability of distribution feeder microgrid and then also throw some sort of a trade-off between these two factors and it's very important to provide an interaction I want the key components of distribution feeder microgrid which are built in research goal ones two and three research goal one we build the power system architecture for the microgrid and then we build off the communication architecture research goal two and then we have the decision-making and control system overlaying control system that just oversees all of the activities to make the correct decisions for the operation of the microgrid so this research goal that is going to provide this interaction among all of the activities that we have on the under the other research goals and then the end goal is to provide a smooth path from modeling the algorithms the these designs that we have in the research goal one to three from modeling to simulation and then finally to an effective deployment and demonstration so that is the main goal of this research goal and then the objectives that we have so the sorry the objectives is to build some sort of a hardware in the loop testing facility at UNO and NMSU Olga we briefly talked about that and also develop realistic scenarios for the operation of the distribution feeder microgrids under various stress conditions and demonstrate the improved resilience and sustainability and please go to next slide so I'm going to quickly introduce the team members at UNO is myself and Professor Jean Ler and then the students that I would like to thank them for their contribution to the microgrid must also microgrid which is one of the main sites for the deployment activities they have done a great job and also the students that have worked on the hardware in the loop testing laboratory to preparing hardware in the loop testing setup next slide please and here you can see the NMSU team they have also done a great job although briefly talked about the activities that they have done for completing their hardware hardware in the loop testing facility and they have done a great job in Southwest technology development institute next slide please so one of the major components for this research goal is the test path so the idea was to choose the research test paths that are actually a good candidate for demonstrating a distribution feeder microgrid and on the right picture you can see the different research sites at the Santa Fe community college we have the microgrid lab and then we have also the Los Alamos circuit 16 at Albuquerque we have the PNM phosphorous site and then the sodium 14 which supplies mesodosol microgrid and then at Los Cruces we have the EP lab and also SWTDI so these are the main test paths that we have in mind for the deployment purposes and also in parallel as I mentioned as one of the major components we want to develop hardware in the loop testing facility and we have ongoing activities at both UNM and NMSU to build up this facility hardware in the loop facility which I'm going to talk about that in the next slide so this is the HIL test thread in UNM so in general the hardware in the loop testing can be categorized into power hardware in the loop and control hardware in the loop power hardware in the loop takes care of integration of the power equipment like for example inverters PV systems or for example battery energy storage systems to the rest of the simulated model in a real-time digital simulator like OpalRT which both UNM and NMSU is using and currently at the mesodosol microgrid we are actually building this HIL lab and the plan is to energize it hopefully by the end of spring we have purchased all the equipment and we're just waiting for some electric job to put the equipment together and then energize it and then the facility is going to be ready for the power hardware in the loop testing and you can see that we have purchased a regenerative grid simulator some inverters a PV simulator and then the inverter control modules which are all integrated into the OpalRT as the real-time digital simulator and next slide please but more recently actually we could develop a control in the loop test thread I would like to thank Vinod and G1 for taking the lead and the reason that we needed this test thread was in research goal one we developed a cyber secure control algorithm for the distribution distributed control of distribution feeder microgrids and we needed to provide some hardware in the loop testing results and we developed this test thread and as you can see we have the Raspberry Pis for Raspberry Pis which hosts the control of each inverter or distributed energy resources and then this is a small microgrid with 4DRs and then these Raspberry Pis are integrated into the actual system which is simulated inside OpalRT the top right photo and you can see some graphs for example when we apply the cyber attack on the frequency control of a microgrid you see that the frequency is fluctuating and we have diverse frequencies for DRs which the end result is just it's going to collapse the microgrid and it's going to damage the microgrid in terms of the frequency stability but when we apply our attack detection and mitigation on the Raspberry Pis we have the stable operation for the microgrid in terms of the frequency and this test thread can be highly used for the research code 2 especially and also research code 3 when they apply the communication and control techniques and they want to see what is their impacts of their algorithms in a hardware in the loop testing test thread. Next slide please. So as a part of research code 4 we have partnered with actually some local industries and vendors in New Mexico and our technology as I mentioned has been a great partner especially because of their recent DC microgrid projects in New Mexico we have been working with them and also PNM especially with their goal to have it carbon free by 2014 they have been a great help they have been collaborating with them and also every ZMS and El Paso Electric also have been a great partners for this project and please go to next slide. So for this year the work plan actually from the UNF side the major goal is to complete the test thread especially the hardware in the loop testing setup and also from anyone's view side complete the SWTDI test thread and enhance it and Olga mentioned about the recent improvements that they had to their hardware in the loop testing facility and Olga and her team are going to host the workshop to define the relevant DfM scenarios it's going to be an October 3rd and 4th and then we are also planning to build the distribution feeder microgrid models based on existing and prototype feeders and also we have already implemented the model of the Studio 14 in Ophalarty in Simulink and then we are also working on the open DSS and then moving forward when the hardware in the loop testing facility is ready we can utilize that and do the actual testing of the algorithms that are developed in the other research goals on a real hardware in the loop testing facility considering the Studio 14 atmosphere also and next slide please and this is going to be my last slide so in terms of the Keras research group coordinations we have been coordinating with the search goal one so you saw some of the results that we developed the algorithms in the search goal of one and then we developed a hardware in the loop testing for the algorithms that were developed in research goal one we are also working with research goal two especially the control in the loop facility that we developed with Raspberry Poise and Ophal can be utilized for testing cybersecurity or the communication techniques that were discussed and also we have been working with the research school three for data sharing and data set acquisition and utilization so with that I hand it back to you and thank you very much thank you Ali so we've purposely left time to answer some questions today we'll make sure to end right at 1 p.m. but I will just work through in the order that they have appeared and the first one is for Manel it's from Adnan who's working with Trilsay and he asks are we also focusing on anomalies generated by extreme weathers or power outages yes well in principle our algorithms they deal with any kind of anomalies and the only constraints that we are applying are related to the distributed you know they probably distribute mission models but we are not limited in terms of the applications of the the applications of our algorithms so the answer is yes I guess great and I think this one's also for you Manel without the reference slide it's a little harder but it's it's from Carl benedict at UNM and he said what is the structure of the objective one activity 1.1 level three data repository content what's the size what are the access methods let me see 1.1 well here we are just starting with that because we still don't have a clear idea what are all the data that we are dealing with so and also I'm probably not the person to answer this but probably my my colleagues would be better than me to to answer that question so what I'm gonna do is to not to leave this question unattended but write a better answer with the help of my group colleagues yeah and it sounds like further discussions that we need to have among all teams so on the same data train and somebody else can pick this up maybe we will start with Olga for all activities producing data how are you documenting those data for use and sharing within your teams between teams and for future discovery and reuse and I'm sure Carl himself would have a webinar he could give on all of this Olga yes so this is a very important point all of the data that we are generating must be available for not only our own researchers but also available to the general public general research community and so as part of this project we have we're basically looking at uh complying with all of the requirements of saving the data preserving the data as Anne mentioned at the beginning there are basically mechanisms for each university to archive each of the data sets and what has been done with this data in terms of results and analysis so what does that actually mean that means that if anybody has a request for any of the data sets that request can come through either EPSCORE or either the universities and basically responsible researcher will respond to that data or a librarian or an archiver will respond with that data in terms of making the general community aware of these data sets well it's basically the steps that we are doing right now through this webinar or through any of the other conference presentations through any of the publications through any of the collaborations with our partners um we obviously don't just you know archive this data and hide them and not tell anyone but we do present this data and we always make sure that we mention that this data is available just like Manel did with the repository just like Anne did with say stating that all of this data is preserved and so on so on so Anne and Manel is there anything you'd like to add no I think that's great any other um panelists want to add to that okay not seeing anybody unmute themselves I'm going to move on to the next question um from a student for Ali Bidram on attack and mitigation graph I see that the grid frequency is stabilized within 10 seconds of attack but there is still a sudden drop in frequency right before stabilization is there any way to mitigate that too this sudden drop might damage our hardware or user's appliances uh no that drop that we have is not uh is not considered sudden drop uh so we have a slight decrease in the frequency after the attack detection and mitigation is applied and we are done from 60 to 56 hertz and so that is okay but if for example it's gonna last for five to ten seconds it's gonna damage the power system equipment and this attack detection has some delays because uh you assume that uh so the attack is applied and takes some time for the attack detection to figure out what has happened and if there is any anomaly and then once the anomaly is detected it's gonna use the mitigation scheme so we have both attack detection and mitigation so that's why we have a small delay before the frequency is restored to six years so I can actually add to that a little bit um that frequency decrease that is shown on the graph just like Ali said such a short term decrease will be acceptable uh to the equipment and to the electric grid and when I say a short term will be acceptable we obviously need to know how short or how long is still acceptable and so there are requirements and uh specifications for what is acceptable and what is not acceptable so basically we consult with all of the standards and that's where we say okay this is acceptable or if something that we model is not acceptable then our next goal is to make sure that we can design a mechanism that will prevent that great thank you well I want to make sure that we end on time so I will just end by answering one question which was can we access this afterwards and the answer is yes all slides and a recording of this webinar will be on our website um you can go to the website at the webinar area and access slides recordings and see the future webinars that we have lined up as a reminder the next one is October 25th at noon Gary Oppendahl he's a very dynamic speaker um if you have additional questions after the webinar feel free to get in touch with me and I can get them answered and we really appreciate everybody's time today so thank you for attending and thank you very much to our presenters