Utility Industry AI Trends - David Hart - Quanta Technology - UAB 2024

thank you everyone good afternoon I don't know about tying everything together Sasha but I'll I'll do my best here guys to to give you a little bit of an update really I'm just going to talk more about um kind of some of the trends and activities we see going on in the industry today and and kind of what's happening out there with with utilities and with our customers so I'm I have a lot of slides here guys um you know I'm kind of a Geeks geek we talked about the Geeks earlier Geeks geek so but I'm going to talk a little bit about the um the the different the the different Pro but oh did we lose the screen oh ah okay well now I'm not going forward did we lose our our controllers ah there we go thank you guys so overview I'm just give you a very brief overview of Quant technology you know just to to inform you you guys a little bit about that I'm going to talk about industry Trends what we see going on in industry and then really I'm just going to talk about some data uh analytics applications so and uh at Quant technology my team we're running a a fair amount of these Analytics project some of these are from other teams at Quantum technology but I can at least give you a flavor for for some of the things kind of going on in the industry and what we're working on and then it C of course any discussion or questions at the end so Quantum technology so Quantum technology our parent number is Quant services so I think most of most of you guys in utility space are familiar with with Quant Services very large the largest EPC company in North America and and again I'm not going to spend too much time here with this other than and say they they are our our parent company so here's the you know just a very ey chart on the number of different companies tied into Quant services so you can see there so Quant technology is headquartered out of Raleigh North Carolina we're the Consulting Group for Quant services so we're about 400 400 uh employees you can see uh we deal a lot in the patents papers studies other things road maps working working with utilities and again you can kind of see our expertise here and and some of the things we deal with and again these are in the slid so I'm not going to spend too much time here guys just for the sake of the meeting what's a little interesting with quanti technology is it is really a technology-driven company right so we're not management Consultants we do a little bit of traditional engineering companies but we don't do like the detailed drawings a lot of the Quantum Services companies do that that type of work we're kind of in between and so we study we do lots and lots of work with the the utilities kind of in that that middle space and you can see here kind of the why you know we really are independent um we have a unique business uh I think and Technical expertise with software and Hardware Solutions uh testing commissioning doing some really Innovative things there so so that's just a very brief introduction kind of to to where I'm coming from and and you know in the introduction SL you mentioned my background for me I worked in protection and control for another years my my PhD was in power system protection so I did that for a while and I I when I graduated school I went to work that's when numerical relays were coming out I did tons of patents in the numerical relay space and then um I wind up going over to the meter division in ABB and all my protection friends were like for God's sakes don't go to meters you know stay with protection and that's when uh we started talking about two-way communication with meters right so we started working on that for a while and actually the phrase Ami came out of the ab Raleigh organization Ron vaa came up with that name not me Ron came up with that in the marketing team the only reason I know is because I wanted to copyright it so luckily trademark so luckily the marketing guys T me out of it but but so now you know I've been in the Consulting now for about 10 years at Quant technology working with the different utilities and really it's um it's a lot of fun to work with the you know the utility customers they're generally you know everyone is is it's a good group of people and they have very good technical problems that need solving and that's really all you can ask for so what are some of the trends we see in the in the industry today so I put significant load growth you know this is we were talking at lunch a little bit about how kind of things have changed right remember a few years ago we were talking about we may not need transmission and loads going down and what are we going to do and now it's just the opposite we need more transmission and loads going up so there you have the data centers uh we were talking there a little bit uh about that you know I think that's still underestimated what that's going to do to the system that's my opinion electrification particularly Fleet electrification is a big issue a big Trend you know uh the frequency and magnitude of major weather events I've I've had this point up here for the past few years bless you here for the last few years and so uh unfortunately guys we have had this terrible situation in in North Carolina you know we do a lot working with the utilities on the west coast for for Wildfire study but I think we're going to see unfortunately you know more of that also and clean energy mandates and so the clean energy mandates are really driving us it's going to really transform the Grid in the next 10 years I I think it's really I think that's one of the key things the what people don't generally talk about we talk about what we know the kind of the trend and where we're going but there's a whole lot we don't know right so how is that going to impact the grid how's the grids going to operate which problems are your customers going to see which problems on transmission are you going to see how do you have data how do you solve those problems when we don't even know all the problems yet right so it's it's a very big uh big issue I think is going to be coming up in the future as we go forward the AI machine learning you know Trends and utilities these just a few of the things you know up here the automated report Generation Um you you know when I first came to Quant technology we worked uh on a study with a a utility they came to us we'd already started looking because of PRC 27 it's a requirement you know you have to review the protection on your transmission periodically but we had actually National Grid Saudi Arabia came to us and they said we want you to review our transmission line settings and we said okay we can do that how many lines transmission lines they said we have 2,000 transmission lines okay so 2,000 normally for a a a protection coordination study you run 10 12 scenarios so we said well what if we automate everything what if we set it up so we wind up running 500 scenarios per transmission line we ran a million fault case scenarios and we automated all that process and we put in all the logic to drive it and know what to analyze it it was an amazing experience that's more on the data management Automation and that's kind of where we started coming in really working with utilities to automate things and do them faster and more precise no one had no protection engineer has ever ran 500 scenarios to check his settings I mean in every utility you go through you would you would find something there that was previously missed and the thing so but I see that you know the report generation the the quality control those types of things obviously will continue generation and load forecasting the data centers the load growth we're already seeing that right we're seeing that with customers people are talking coming in you know where do we see where do you think this is going to go where are we going there and obviously the the the load forecasting itself right situational awareness this is kind of back to my earlier point of you know what do we not know there's certain things we know so if you know something it's easy to plan for it's easy to track is easy to see what do you do when you don't know right and so there's there's some as things you can monitor today but you one thing to very important to remember in the utility space if you're thinking you're going to be somewhere in 10 years that's not a lot of time in utility right we've all worked in utilities if you're going to install something rotate something upgrade something you're going out to Brown Fields you're going to do certain things in the field 10 years is not a long time to do it so but we see that and so you see all of this capital projects going on in the utility space today you know and and so there's a lot of things going on there and a lot of things we can do to improve situational awareness so what are what are some of the projects that that we're working on today someone asked earlier you know what were the inputs for the grid Edge devices I love that question so you you know the there's a lot of things going on here in in the field today the digital substation the i6 1850 base substation it's a big area I I've been talking about that since my ABV days and I can tell you guys I've seen more activity in the last year than I've seen in my career in the 61850 space so it's the volume 61850 is driven by volume but once you move to 61850 and you really are going to a digital substation there's so much more you can do on the analytics side right that doesn't happen overnight because you've got to roll out your 618 50 and you've got to get enough critical mass to do that right but if you plan it it can happen so 61850 is definitely one of the inputs you know the field sensors there's these smart fcis out I don't know if you guys are familiar with those those are pretty neat neat Technologies uh Smart Meters always smart meters are just a amazing thing I think with the Ami 1.0 you know when we first launched Ami the big issue was how do we interface with it and how do we get them tied in the system and there's just tons of discussion and effort around that that's all pretty much worked out now everyone's done that Ami 2.0 is really the next step of where I think we would have liked to have been with 1.0 and it's just as different right so the the things you can do in the meter today have nothing to do with the meter Department right and that's one of the interesting things with analytics the other stuff it's all of these cross uh cross area knowledge you really have to pull together to try to figure out how to leverage this if you go use your Ami 1.0 spec for your Ami 2.0 system you're going to get

a 1.0 system right so but and all of these are reaching end of life all these early Ami systems you can see that moving in the space PQ meters scada data GIS data all this has been around for quite a while you guys were talking about the imagery lar in particular I think with the drones tremendous opportunity there on the on the inspections and the other things so the output there's just lots of things you can you can do once you put this type infrastructure in uh in place one like fault location you know particular on transmission you can you can narrow it down to the tower the event analysis protection you just look for where people spend their time protection Engineers spend a lot of their time doing event analysis when really there's you know 90% of that can be automated so there's a a lot of different things you can do here in in terms in in the space and and what you Monitor and how you go go forward grid situational awareness you know guys I've seen um couple utilities really focusing on this because they know that there going to be issues on the grid they haven't seen before and there's one case we did a paper with a utility but they they had done this uh earlier but they were talking about a customer had a a flicker issue right and turns out that the flicker issue was caused by two large ivrs right and interaction between the controls without monitoring on the transmission grid they never would have been able to figure that out so even even when the customer notified them they had to go dig through and look for it right that took a lot of planning that took a lot of expertise they had a great system in place they set up to figure out to go through and analyze that so these are just some of the other things uh you you know you see coming up asset uh monitoring Asset Management obviously is a tremendous area that could benefit from Communications to grid Edge devices so of another another large area there so here's just a few of the projects that that we've been working on this one we we actually completed it was analyzing uh pmu data is everyone here familiar with pmu when I say pmu data GPS based synchr phasers U and the GPS timestamp synchr phasers on the power system so so this was actually a doe project that we work with you know with the the the doe and we we went through and so this this was a a project really to look at kind of how you can analyze ton terabytes of pmu data so the problem with pmu data is not having the data it's knowing when to go look at the data right so that's the problem we were focused on in this project and we did a lot of things to speed up the processing and look at the data and what you can do we went through and came up with this this new algorithm for event detection to kind of highlight where you need to go where you need to go look what you need to see so this is a a new technique we we came up with and we're actually looking to to leverage this in a a subsequent project that that we're working on and so we we that's an algorithm we came up with we're in the process of patenting that and going through uh accurate uh location this is a project we did with pg& we've been working on for quite a while now on the analytics front this is kind of our first uh first project we got analyzing data from the transmission substation and so what what you do is you you can extract information from the from the IEDs uh in the substation and pull that back right and you can figure out you can see the graph there where you you oh I lost the image the graphic here where you're showing you can kind of map to the Tower uh the nearest tower for the fault location so it it's a great thing it takes in different types of data runs different types of algorithms tries to give you the option of the best algorithm and and the location and the which algorithm you pick is driven by kind of data sources available and other other things like that so this is this is a project we we've been working on uh we're still working on with with these guys actually we're doing some some additional things looking there oh I lost my go forward ah okay here we go so here here's another another project uh goat the goat project as I refer to it right so this project actually we're just getting ready to start uh we should start it kick it off this month this this is a project to look at um interactive dashboard for kind of looking at the health of the system and this is a little bit I think on The Cutting Edge because you're looking with system operators and how you would present this type of this type of information but we we're working with you can see here the group of utilities the National Lab uh we we put this this project together um so the idea really here is to pull in pmu data from across the transmission grid look at that and try to help figure out more in real time the status of the grid so there's there's a lot of details here but you can you can see there but that's that's the idea and so this is kind of focused on the on the transmission grid and really improving the situational Awareness on the transmission grid so another project switching over to distribution we worked on with customers we did some cluster analysis for basically for for voltage variations across the distribution grid trying to identify if it's an individual case or a systemic case and work with customers to go through and and map that out that's that's something we've done in the past this was done with 1.0 Ami 1.0 data right and again I would expect a lot of improvements with what we can do there with the 2.0 data the another package we're working on with utilities uh using the Ami data to do some some basic things here uh Transformer service Transformer Asset Management outage reliability metrics grid modeling power quality analysis anomaly detection meter health manage again this is really kind of the first step to leverage the 2.0

functionality uh in the system and I think this is going to be a a very large area as we go forward right so the the trick with the 2.0 meters is to figure out what you want to do with the 2.0 meters they have a lot of power a lot of processing power a lot of capability but how do you leverage that right how do you tie that into your processes David yeah I want to take a second to Define what's the difference between Ami 1.0 and 2.0 well 1.0 I design so it's

old 1.0 1.0 we came out with around in 2000 and really you know all the there were four like four major players at the time and but in essence everyone's system was really basically the same right it was very focused on um the energy usage and the interval data and some instrumentation data and pulling that back and that's that was a very standard thing uh we did it was a a great project I think it really pushed the the industry forward it's the first time we've ever had sensors across our distribution grid and that's what I kind of call the Gen one the the Ami Gen 2 has a slightly different focus in that really the the meter itself is they put a lot more horsepower in the meter right so for example one thing you can do with of the the 2.0 meters you can pull back subcycle data and that's that's we we could never do that before in the 1.0 right even with a chipset it wouldn't support it so so now you can pull back waveforms there's certain things in there most of the guys have things like phase detection in the 2.0 meters which is really important you know distribution forever is operated on rules of thumbs and averages right yeah yeah yeah feel like I'm at church amen so but the the um the the rules of thumb and averages and and so what things like the phase detection and some of the other smarts are going to allow you to do is get rid of those rules of thumbs and averages you'll be able to improve your your load models you'll be able to improve your grid models and so you can improve your forecasting you can improve your load balancing there's a lot of things that come together with 2.0 I

think we barely scratched the surface there with what we're going to be able to do you know the old old systems like um uh the ABB system when they used to do state estimation on distribution they had skate at the substation and then they allocated the load by the Transformer rating down all the feeders you know it worked right but nobody knew if it were correct right and a lot is true to with the Ami analytics today on average it's true right so because you take the skate at the head in and you use the load data and you balance said and you you say okay it it looks correct and it is on average right Engineers we hate averages right we want to know the status of what's going on because it's the individual feeders or where you're installing a a renewable site those are the things that are really important to know that specific circuit and what's going on so I think we're going to get a lot better at how we do that uh going forward in the future uh distribution system resiliency so someone you know we're talking about extreme weather events you know how how do we optimize that and look at that and kind of figure out more than the reliability more of a resilience approach right how do you harden the system where do you want to place sectionalizers reclosers and other things and kind of automate that process you're always looking for some sort of Matrix with some sort of waiting Factor right so the utility can go through and figure out okay if I'm going to prioritize for this goal what what's the what's the different features and how do I prioritize and how do I do that so we've been doing quite a bit of work in in that area again the recloser placement uh you you can see here we've been going through and automating a lot of that pulling it together with the and even suggesting sites for tie ends for to improve resiliency reliability those types of things again a lot of analytics to improve the process and kind of go beyond what we done today Wildfire risk analytics this this is probably one of the widest topics I've seen in quite a while you you know there's a really for my protection background there's a lot of things to do in protection related to Wildfire analytics but you also have weather analysis you know you have uh there's a ton of variables here you need to look at so the Wildfire risk exposure identification benchmarking everyone now is being asked because you would think most utilities are getting the question from people what is what is your Wildfire plan what is your your mitigation plan well no one's really come up with a wildfire Plan before right that was kind of a West Coast issue West Coast has done a lot on that but now you see that moving across uh you know a lot more people are looking at that and what do you do and how do you try to mitigate that one thing we did this was a project we did with San Diego and it's this is actually more of a control than analytics but but one problem we we tried to solve with these guys early on we were looking and saying well what if your your distribution conductor is broken and falls to the ground at arcs of course that could obviously start a fire right so we said well what what could you do in that if if um what could you do to to mitigate that risk well we we work with these guys and came up with a a an interesting idea and we called it falling conductor not it's actually not falling it's falling conductor because a conductor takes so many milliseconds to hit the ground I think it's 40 milliseconds if you drop it so if if the conductor breaks and falling on the ground you can actually detect it and trick the recloser or breaker before it hits the ground a great idea we we patented it uh or sg& patented it they we work with them they patented the idea and so they've been using it out uh and on some of their their feeders going through that that was a great project to work on um we've also been working uh with different companies you know on the to look more at the approach of what they need to do like pg& and the CPU we're looking there optimal operational strategies for PSPs that's the public safety power shut off you know if they have very dry conditions windy conditions then they'll de-energize some feeders there's just a ton of protection and analytics you see rank the circuits manage load generation packs all this ties together and that's why I say I think Wildfire is a very interesting area technically right because it does tie together so so many different aspects uh and and types of expertise uh Wildfire mitigation program design and implementation you know we've been working with with Luma to look and go through and kind of assess identify the risk assess the the risk management review your design standards processes system level process protocols and rate Wi-Fi risk so so these types of things again this is pretty standard as you start figuring out kind of you want to move into this area and figure out how you address this and another similar project worked with heo on on a project very similar to the one we we've working on with PG electrification um you know electrification is is definitely going to be we've worked with a lot of different companies here on this uh looking at road maps and how do you do and how do you roll that out where do you need to put charging stations there's a lot of details a lot of things here we've done a lot of work with micro grids right so you you can see all this to kind of come together this this is a big area everyone's obviously more concerned about the fleet electrification because of the larger impact on the grid and what they're doing so again there's a a lot of activity going on in in this space and then uh you know kind of you could go through a standard process here again to to look at it and all the analytics kind of pull into this guys so the bottom level up you look at your your locations right forecast adoption loads map and aggregate the facilities this we've actually been working with utilities kind of forecasting the the load um and again I think utilities have been been pretty conservative on that and so I I I think there's going to be a lot more coming coming coming out uh than we see today so uh another just another example of utilities we work with for electrification so um you know but looking at Fleet loads and how would you approach that I'd already mentioned the fleet loads how you can go through and look at this you know how do you size the load load's not uniformly distributed so they were originally trying to look kind of uniform distributions it's very site specific so we can look at the manage charging you know if you can manage the charging of the fleet versus just having everything done randomly right and then you look at the profiles and go through but but again another another project here lots of information a lot of data lot of analysis on how to do that and I really didn't touch on data centers guys I had a slide but I took it out because I knew I was going to talk for too long but I I do think the the data centers you know we're working with folks I I mentioned the the forecasting and the sighting and the the also starting to look at some of the monitoring aspects of what you can do there the loads are so large and so impactful to the grid you know I I think that's going to Warrant additional additional work and effort there so that was a very quick overview guys I apologize for running through the topic so fast so but I really just kind of want to give a flavor overview of what we're seeing in the market on the analytics front so any questions one over here David so uh I'm Daryl I want to introduce myself I'm one of your uh Partners we work with uh Quant uh Technologies um we work with deir and Carl Wilkins oh yeah yeah but what I wanted to see if you had an opportunity to talk about is some of the to tools that Quantum technology uses around lvd and the reason that that is important from a data perspective is as the utilities are really trying to get to zero carbon by 2030 and 2050 is mandated by the PSC there also this pressure coming from the infrastructure Bill and nevi as it relates to EV and Equity as well and really the figure out how Quant has created Tools around locational value de to really support the utilities and their whole strategy around nevi and EV in communities where there's some challenges around equity and where you put EV and how you get to zero carbon by 2030 or 2050 in the communities yeah no thank you I and we we have done exactly that I think you saw we've done quite a bit of work and working with people to figure out the the sighting and the the the interface and kind of the planning and how you tie it all together we've created tools with that Carl's actually driven a lot of that work and a lot of those initiatives on the on the distribution side so I I think Carl would probably be give a much better detailed answer to that than I would but I know he's done quite an extensive lot of work in that in that area yeah and some of the data collection of that is is extremely important as the youtil ities look at what their public private partnership is going to look like under nevi and getting the federal funds from the infrastructure Bill using that data to be able to say what their EV infrastructure is going to look like right so there that that's where there's some some leverage data points around that as well right hey there uh you shown a lot of examp examples of showing data analytics for the utilities to use have you worked with any utilities on customer facing dashboards I mean a lot of what we're trying to do is educate our our users on why rates are what they are why and try to to educate our users on some of that have you seen any analytics that could pushed out over portals to customers on on electric usage well I mean obviously all the Ami systems have portals that go out to customers today so you know provide information on usage and there's some different rate comparisons and different things you could look at that that's pretty common uh in the industry today was there something beyond that specifically you were thinking or especially when we're we're trying to see ev adoption and that sort of thing is there any type of tools that they could see can my Transformer handle it you know so stuff like that that I've seen that uh what what impact would would solar panels on my roof versus an EV at my house do for me and and that sort of thing is there anything that that utilities are creating in that respect no you know I haven't really seen that's actually a great a great thought but no I've not seen any any any kind of tools there and I I think the the bigger thing is you know most customers are assuming you're going to upgrade the Transformer right so that's really more of a a utility issue how how do you address that and how do you you you fix that that that's actually one of the big things about the service Transformers uh in in the Ami and again that's been around for a while don't don't get me wrong there's a lot of work been done on the the service Transformers and Ami and other examples but 2.0 will offer some some vast improvements uh over what's been done previously it's going to take a lot of the guess work out so so I'm sorry but to answer your question I I think no I haven't really seen that type of request from from utility or I'm not aware of it Carl Carl may be aware of it Carl Wilkins another gentleman in our our team but I'm not question for you um what Don's talking about kind of ties in with the Ami 2.0 um architecture but part of that architecture is that uh third Pary should be able to develop applications to run at the meter on the meter right um are you seeing that opening up because right now I think most most of the time they're developed you have to use the SDK of the meter manufacturer is that is that changing have you seen you know we just got the SDK so I I do see that kind of kind of opening up um to where you can do more applications now I will say you know some of the things we talked about the the phase detection and other things that those are really those those apps that that are coming out with the meters uh out on the market today so they've already included some in there you have the ability to do more in there and actually I think that's going to be one of the large things to figure out here now with 2.0 as we take the next step is how do we leverage that and what do we want to leverage there there's some things there like interface to to home uh you know the solar panels on the homes and other things they talk about for control I'm I'm a little skeptical there just because I you know when we did the the first round of meters and we were trying inhome thermostats and other things you run into all sorts of propagation issues that aren't easy to solve right so that's just me personally that doesn't mean they haven't overcome it somehow personally a little bit of skeptic there but I think there are other areas we really do need to look at and see uh you know how we can leverage that type technology

2024-11-17 18:06

Other news