Jeff Freedman, a research associate at UAlbany’s Atmospheric Sciences Research Center (ASRC), is working in collaboration with researchers and meteorologists to create the Wind Extremes Forecast System (WEFS). By combining numerical weather prediction (NWP) modeling with machine learning techniques to produce forecasts, the system will seek to minimize power outage impacts for millions of New Yorkers. Freedman shares on this episode of the UAlbany News Podcast how WEFS will help utility companies and emergency managers mobilize resources more efficiently, reduce restoration time and improve the long-term resiliency of the state’s power distribution system.
Jeff Freedman, a research associate at UAlbany’s Atmospheric Sciences Research Center (ASRC), is working in collaboration with researchers and meteorologists at the Consolidated Edison Company of New York (Con Ed) and MESO, Inc. to create the Wind Extremes Forecast System (WEFS).
By combining numerical weather prediction (NWP) modeling with machine learning techniques, the system produces real-time wind speed and gusts forecasts in an effort to minimize power outage impacts for millions of New Yorkers. The project is sponsored by the New York State Research and Development Authority (NYSERDA)’s Smart Grid program and will use data from UAlbany’s NYS Mesonet, a statewide network of 126 weather stations.
Freedman shares on this episode of the UAlbany News Podcast how WEFS will help utility companies and emergency managers mobilize resources more efficiently, reduce restoration time and improve the long-term resiliency of the state’s power distribution system.
Read more on the project: www.albany.edu/news/91213.php
The UAlbany News Podcast is hosted and produced by Sarah O'Carroll, a Communications Specialist at the University at Albany, State University of New York, with production assistance by Patrick Dodson and Scott Freedman.
Have a comment or question about one of our episodes? You can email us at mediarelations@albany.edu, and you can find us on Twitter @UAlbanyNews.
Sarah O'Carroll:
Welcome to the UAlbany News podcast. I'm your host Sarah O'Carroll.
Sarah O'Carroll:
I have with me Jeff Freedman, a research associate for UAlbany's Atmospheric Sciences Research Center. He is collaborating with researchers and meteorologists to reduce the risk of power outages in New York state from extreme wind events.
Sarah O'Carroll:
So you're leading a team to create a wind extremes forecast system or a WEFS. What exactly is the system designed to detect and predict?
Jeff Freedman:
The system is designed to predict extreme winds or excessive winds and wind gusts over the particular region that one of our partners, the Consolidated Edison Company Of New York, oversees. They're a utility and they are interested in predicting power outages and developing a system where they can more accurately predict these outages, which would allow them to position crews ahead of time and save some money in terms of the amount of people they have to have in the field and shortening the amount of time that the power is out.
Sarah O'Carroll:
Can you break down a little bit how it works and what kinds of techniques that you're using as far as the technology?
Jeff Freedman:
The project, it's funded by the New York State Energy Research And Development Authority.
Sarah O'Carroll:
NYSERDA.
Jeff Freedman:
NYSERDA. Has a couple of partners I mentioned Con Ed and our other partner is MESO, Inc. They are a atmospheric sciences firm located in Troy, New York and they'll be working with us at the Atmospheric Sciences Research Center as part of the University at Albany to develop a forecast system. The system is going to incorporate numerical weather prediction. Those are the weather forecast models that everybody's familiar with in terms of getting their forecast from TV and radio or from the internet.
Jeff Freedman:
And then we're going to combine that with information observations from the New York state Mesonet, and that would include the surface observations as well as a profiler. Instruments in the profilers measure the temperature and humidity and the winds up to several kilometers or a few thousand feet into the atmosphere. With that information, we're then going to use what we call artificial intelligence or machine learning language to make adjustments to the forecast and more accurately depict what the winds will be on very local scales.
Sarah O'Carroll:
And what would you say is the threshold for predicting a potential power outage? I just want to get a better sense of how intense we're talking in terms of sustained winds and gusts and any other factors that would potentially create an outage.
Jeff Freedman:
The first thresholds we'll be using are provided for us by Con Ed, and there we are looking at the wind speeds of 30 miles per hour and 50 miles per hour as to two thresholds. That will be sustained winds, so those are winds that are sustained over a period of time, whether it's minutes up to several hours. And also wind gusts, wind gusts are typically five second maximum wind speeds over a particular period of time. Those two thresholds we can use because depending upon what kind of weather you're dealing with will also depend upon the type of extreme event that may cause a power outage.
Jeff Freedman:
So, for instance, if you have a winter time storm, a snow storm with heavy wet snow that tends to cling to the trees, lower wind speeds will be more effective at bringing down trees and tree limbs than otherwise would be the case if there was no snow. Also, if you have trees that are in full leaf, you also have to deal with the fact that there's more pressure, there's more force that's being exerted on a surface area for a tree, and so therefore that tree is sustaining more of the force at the lower wind speeds than if the tree leaves were not on the tree. So we look at these two different thresholds.
Jeff Freedman:
And finally, if the soil, the ground is saturated, the tree roots tend to be a little more weaker. Right? Soil is loose, and so therefore trees would tend to be more susceptible to being downed at the lower wind speeds, let's say the 30 to 35 mile per hour winds instead of 50 mile per hour winds.
Sarah O'Carroll:
How would you say this technology will serve power companies as well as emergency managers?
Jeff Freedman:
Right? So Con Ed and other utilities need to have an idea of when they're going to have an extreme event or when they're going to expect some power outages. The more confidence that they have there's going to be an event happening, then the more confidence they have to put crews into the field, to position crews a certain places. It's a question of not only timing but also location. So it's one thing if you know there's going to be very high winds over a period of hours, perhaps a couple hours ahead of time, perhaps a day ahead or even up to two days ahead, we'll be forecasting out to 48 hours. But also the location.
Jeff Freedman:
A very strong aspect of this forecast system is the ability to locate where we think the extreme winds are going to happen with an area of less than, let's say, a few thousand feet or a mile, really using a kilometer in terms of what the model resolution is. So being able to position those crews is very important. And also being able to, once you have the crews positioned, then that reduces the amount of time that the power is out.
Sarah O'Carroll:
And are you looking at a particular area of New York, or is this designed to be statewide? Or do you have a perhaps pilot area that you're experimenting on right now?
Jeff Freedman:
We're focusing right now on the Con Ed service area, and that includes New York City, Westchester County, Rockland County, parts of Orange and Putnam County as well as parts of Sullivan County, and a little, I think a couple of sections in Northern New Jersey. Of course once the project is done, then we can see if it's ... It'll certainly be exportable for other utilities to use throughout New York state and beyond.
Sarah O'Carroll:
So we've talked about the utility companies and the other people who might be able to use this, but how might atmospheric scientists build upon this research?
Jeff Freedman:
Well, any time you develop a model of this sophistication to look at a particular parameters such as winds, you're going to learn something. So we're going to look to see what the best combination of model physics is, how to best assimilate the New York state Mesonet data, how accurate these kinds of forecasts are compared to, let's say, existing operational forecasts. This will tell us what kind of improvements we're making in basic forecasting particular parameter. But we could also look at other parameters that are involved in making these kinds of forecast. So we can compare this with other operational forecasts, high resolution forecasts that, let's say, the National Weather Service uses. Then when we publish information on this, people can look at this as a lesson learned and say, "Hey, you know what? This is something we can use to improve operational forecasts."
Sarah O'Carroll:
And what's the timeline for this project?
Jeff Freedman:
Well, right now we're in the middle of a two year project. So we're finished testing the model in terms of it's set up and the types of information that we're going to be bringing into the model. We're going to go into what we call an operational testing phase. This is where we actually run the model in real time. We'll be doing that for the next six months and this is how we'll be able to determine how well the model performs. After that, we'll have a few months where we review, work with Con Edison and the [inaudible 00:07:16], how did it perform in terms of predicting where we expected load outages, power outages to be compared to what your existing outage models were showing. And then we'll have a final report. We'll wrap that up with NYSERDA and then we'll look to see where we can go in terms of other utilities throughout in the state or beyond.
Sarah O'Carroll:
What are some trends in extreme weather events in New York state including wind events that you'd want to speak to and that this technology would help us better understand?
Jeff Freedman:
When we put together this proposal to NYSERDA, the first thing that occurred to me was, is there any kind of trend in extreme wind events in New York state? And we took a look at the storm database that's available from the National Climatic Data Center and we discovered that yes, certainly there has been a a definite increase in the amount of extreme wind events throughout New York state, and there could be associated with any types of weather. It can be associated with thunderstorms, winter storms, tropical cyclones. So we do see a general increase in these events that have been leading to damage, power outages and other types of injuries even to the population. So that trend has been quite discernible over the last 20 years.
Sarah O'Carroll:
How would you say this technology fits into Governor Cuomo's Green New Deal? I know an emphasis was about innovative technologies.
Jeff Freedman:
In terms of of having more wind energy on the grid or more renewables on the grid, we have to deal with, is there going to be effect on transmission and since we can now have a more accurate forecast on that, we could say, all right. You know, it might be a question where we have to move energy from this part of the state to that part of the state, where we have to make sure that there are crews in a part of a state that's receiving transmission from, say, wind farms upstate or wind farms even off shore. If you have a better transmission system where you reduce the amount of outages that you have or the length of time of these outages, then that really facilitates the bringing on of renewables onto the grid. And it also has to do with the fact that there is also a plan to establish micro grids.
Jeff Freedman:
Okay. In other words, these are sort of what you consider not quite, but almost islands of generation and they're mostly going to be renewable. So, in other words, universities can be thought of as having a micro grid where they have power that's being generated on campus or near campus. And if you know that there is going to, an event can be happening, well, you make sure that micro grid is up and running and has enough power to supply essential services, let's say, to a university or to emergency management services in general. So we're looking ... In the longer term and you'll see more of these micro grids established, and having a better idea if there's an extreme event coming means well, we know that these micro grids will be able to supply power during that.
Sarah O'Carroll:
Jeff, thank you so much.
Jeff Freedman:
You're welcome.
Sarah O'Carroll:
Thank you for listening to the UAlbany News podcast. I'm your host Sarah O'Carroll, and that was Jeff Friedman, a research associate for UAlbany's Atmospheric Sciences Research Center.
Sarah O'Carroll:
If you find this kind of research fascinating, you might enjoy an episode from last season with Jerry Brotzge, the program manager for the New York state Mesonet at UAlbany. You can let us know what you thought of this episode by emailing us at mediarelations@albany.edu or you can find us on Twitter, @UAlbanyNews.