[Future of Work Ep. 6] The Future of Renewable Energy with Molly Bales

When it comes to the renewable energy industry, it is crucial for developers to have access to clean, relevant sets of data for them to be able to use data effectively and optimize revenues for storage.
blog image

Article continues below

Episode highlights:

  • When it comes to the renewable energy industry, it is crucial for developers to have access to clean, relevant sets of data for them to be able to use data effectively and optimize revenues for storage. Multiple sources of raw data are available, from in-house sources or from independent system operators, but not all of them are correct or standardized.
  • To be able to use data to its full extent, professionals in this industry need to be able to engage with data intelligently and to be critical when it comes to data sources, as legacy systems are still being used extensively.
  • Artificial Intelligence is a promising technology when it comes to the renewable energy industry. In the future, AI will be responsible for important decisions that will lead to a sustainable source of energy, thus being able to incorporate storage into the grid - a crucial step for unleashing renewable energy.
  • Since setting up a new facility is a project that requires up to four years to be completed, using current data to forecast the outcome is mandatory, as this can give an important glimpse of the future facility’s success and profitability.

 

Transcript (edited for clarity)

Claudia: What is the last thing that you learned that got you excited and why?

Molly: Recently, in Texas, the electric grid suffered a major setback, and we had major outages during February, it was a very interesting time to be a member of the renewable energy industry while all of that went down. There's been a lot of industry discussion, a lot of policy discussion at the state level about what do we do about this? How do we prevent this level of outages from happening in the future? There are a lot of different things that can be addressed. One thing that I've found interesting is that a battery can perform lots of different grid services. As a developer, we must think about what is this battery going to do. What is it useful for? And then think about what is that value and is that value recognized in the market? Can I make money from it as I am building the project? One potential revenue streams that is particularly exciting, and that doesn't come up quite that often yet is called "black start". With black start, if the grid goes completely down, you have to help the plants that need to get everybody back online, by supplying them with the energy they need to power on.  If you have a major outage across a huge portion of the grid, you end up with this problem - how do you start up the plants that have to get the whole grid going? There are these special black start generators that are responsible for that. In the event of a huge grid outage, those are the responsible facilities. Something interesting on the storage front is now is that we're reconsidering the possibility for batteries to provide black starts. If the entire grid goes down, it could be batteries that bring the other generators up to speed and end up restarting the entire grid after a massive grid outage. So, I thought that was pretty exciting!

 

Claudia: In terms of renewable energy, how central is data to your work and more importantly, how well do you think it's used in the industry overall?

Molly: The use of data flows through the entire development process and operations process, and there are some ways that we're good at data, and there are other ways that we're not. Not surprisingly, I think as developers, we are always on the hunt for the best data we can find. We do run into issues where, unfortunately, sometimes the data sources that we have are just not strong enough. We are making a massive change in terms of how we run the grid and we are, as developers and operators of renewable energy, looking for way more granular levels of data. We want to get down to the electric node level of understanding the pricing differences what's happening at these very specific points throughout the entire grid. Sometimes it's hard to get those pinpoints of data correct. It's also can be difficult even figuring out where can we find some things on a map, the data that you try to pull is going to be a lot of times pretty messy and it's not standardized. There's a lot of room for improvement there in terms of the raw data sources that we have to pull from. So I do hope that improves over time. I work a lot on how we contract our facilities and how we get paid for the energy that we're producing. And for that we are scouring for all types of data, trying to understand what is it worth to someone to contract with us for this energy, but also what is it worth if we decide to just go to the liquid market and go to one of the independent system operators – ISO, those independent system operators which set up these liquid markets. For that, you have to have a great understanding of what is going on in the market because you are an active energy trader if you decide to play on the market directly. In storage work, in particular, the data becomes much more crucial because there are a variety of grid services that storage can provide, but a battery can't necessarily do them all at once. You have to decide - do I want to charge my battery right now? Do I want to discharge my battery? If I'm discharging my battery, what grid service am I performing? And there are a variety of different ones you would do, you need to know what's your best option. So it is crucial to understand that data is incredibly important not only for optimizing revenues for storage, but also for any energy facility that is playing in the liquid energy market. 

 

C: What's the future of work in renewable energy?

M: I think it's been kind of interesting in terms of our industry that, as I said, we are growing, I do think there's some consolidation going on. That's kind of an interesting trend, for better or for worse. I think that, on the data side, there are certainly startups and people doing kind of interesting analytical things, but there is still room for tools, for us to do our work better. That's definitely changing and heading forward. I think a lot of the trends that we see in our industry are similar to other industries. I mean, certainly, post COVID and actually even pre COVID, this is an industry that's very comfortable with remote work. If you're a developer, you're going on to customer sites all the time, you need to stay in touch with people while you're on the road. So it's just natural that we have people who work remotely, who aren't necessarily in one of our main offices and that's perfectly okay because you have to trust your teammates to get their work done while they're on the road. Why wouldn't you trust them working from home? I think with COVID obviously things changed dramatically and people got even more out of the office and I think that's a big change and I don't see us going fully back.

 

C: How do you think the renewable energy industry could use data better? What are some obstacles that are standing in the way of the industry using its data assets better or more efficiently?

M: I think I may have alluded to this earlier – the problem is getting the data sources standardized and getting them to be correct. In some cases, believe it or not, we have to use Power Flow software to figure out where a substation even is. Power Flow Software is a way of understanding how electrons are flowing throughout the grid. Some of my colleagues use this so they can figure out and trace the place for substation, for instance, but for those of us who are in development based on the data sources that we may be getting directly from the market itself, the ISO, we may not be able to figure out where it is. That seems so simple, right - just tell me where the substation is - but apparently the coordinates are not correct or something has happened where it isn't correct and you can't always get really scrubbed data. Then, there's definitely a problem when it comes to nomenclature, calling something, usually a node or a substation, by a different name, that's very confusing. In the future, being able to clean up those data sets and give people who are developing facilities clean enough data to work with would be super helpful. Then you can get deep into the optimization of facilities, that's something that plenty of PhDs are working on that. Some of the low-hanging fruit is just clean data, which sounds like it might be a small asset, but it's huge! It's moving a lot of bureaucratic institutions in a way that can be challenging.

 

C: Is there any AI or machine learning being used in renewable energy and storage? If not, what's on the roadmap? What can we expect to see in the next 5 to 10 years? 

M: AI can be used for deciding in real-time - within minutes if not seconds - what function of the battery should be performing and what maximizes revenue and presumably value for everyone who wants to benefit from energy. I think that's a huge area where those quick calculations, that ability to understand what is going to be optimal and continue to make those decisions and run plants better is a huge part of having a sustainable grid and being able to incorporate storage into the grid, which is a crucial piece for unleashing renewable energy. Some startups are working on that kind of thing. It's understandably more difficult for incumbents to kind of have those skills in-house, so we usually see them coming out in younger companies that are grappling with these problems. I think it takes a little while in that, frankly, these are huge facilities and they cost a lot. You always want to put debt on those facilities. That means that you need to get lenders comfortable with how the facility might need to be operated and that may include a young company that has done a brilliant job optimizing revenue stream, but it doesn't have any credit. Right? You can end up in a situation where it takes a little while to get those algorithms implemented in a useful way. But I do think fortunately there are owner-operators and some of them are doing work off of their balance sheets. If they don't require outside finances, then they can get away with implementing new things. In a lot of cases, that's where we see a lot of innovation being incorporated. Long story short, I would say it's happening, those technologies are being implemented and I think it's just getting better and better and it needs more time to establish credibility. I think, over the next five to 10 years, we'll certainly see AI be implemented in this area, too.

 

C: How important is it for professionals to be data-savvy and to be able to use data to its full extent when it comes to the renewable energy industry?

M: I think it's crucially important. There are certain roles where you could get away with not being data-savvy, but there are other areas like my role where you need to be able to take a critical eye to data sources. We have talked about taking a critical eye to various consultant reports, you need to understand what biases are there and then, you need to drill in and see - are these analyses helpful? Do they need to be a little bit more reconciled in such a way that we can gain some insights from them? I do that a lot in my day-to-day work and I think most people in our organization do have to do that as well. I would say certainly say that being critical and being able to engage with the data intelligently is mandatory. You can crunch numbers in Excel and it's great for complex formulas, but doing sanity checks and understanding what you're looking at, being able to reconcile various data sources and piece together, what you think is going on, and getting a coherent story in your head about a market is very important. It takes a long time to build these facilities, and that’s why it is important to think ahead and try and forecast as much as possible. The smarter forecast you can have, obviously the better and the better off your products are going to be. All you can do is look at the current and historical data and then try to project forward and try to guess the future to figure out whether you can build a facility that makes sense.

 

C:  Is there anything that you would like to share with us? What's a cool project or initiative that you've been working on that you'd like to share with us?

M: I work on the storage team, so I'm biased, but I do think that what has been interesting for Recurrent and actually for the entire energy industry is that we are seeing storage being more incorporated into the industry. My company is also very active in terms of storage development and we're seeing that across the board, storage is starting to make sense in markets where it didn't before. California had a duck curve and it made sense to do energy storage in California. And now we're going farther east to be able to implement storage which is exciting. In the case of the duck curve, the more solar you have, the more you see this decline in prices in the middle of the day. What's happening is that there is – for example, if you look at how solar is produced over the course of the day - you get your maximum output of production from solar facilities in the middle of the day. When you have some solar penetration, but not a lot that still works fine, there's not a huge impact, but the demand and supply curve might be relatively evenly matched. What happens is when you have an increase in production during those peak hours, you end up with this mismatch between supply and demand, as there's a lot of energy available from solar facilities and the demand has stayed the same. It ends up dipping in the middle of the day and creating a duck curve. In a lot of cases, if it's a four-hour battery or even longer, you can store the energy during those times when energy prices are super low, and then you discharge, so you let your energy onto the grid at a later time when it's more valuable. You'll see that, over time, as you increase the amount of storage in a market that has a duck curve, it will alleviate that duck curve and make the market more efficient.