• Alex's LinkedIn
• Wattch's Website
• Wattch's LinkedIn

Note: This transcript was auto-generated and may contain minor errors. 

Sean Swentek: Hello, and welcome back to another episode of the Future of CRE Sustainability. I’m your host, Sean Swentek. Today I am joined by Alex Snedeker, General Manager at Wattch. Alex, thanks so much for joining me today.

Alex Snedeker: Yeah, thanks for having me, Sean.

Sean Swentek: I’m excited to have you on the show. I’ve been hearing a lot about Wattch in the market, so excited to hear what you’re working on on both the hardware and software side of things and uh, in renewables.

Talk to me about Wattch’s capabilities. There’s obviously traditional methods of assessing solar site performance, which can often fall short. What have you developed at Wattch to address these limitations?

Alex Snedeker: Great question. There’s a couple of different ways that we like to think about addressing issues that might occur on site.

The first would be alerts, and the second would be KPIs or performance metrics. So some of the issues that we’ve seen with alerting, essentially it’s just having a very low signal to noise ratio, and that shows up in a couple different ways, right? Uh, so there’s a lot of. False positives and false negatives.

So that could mean that, you know, an inverter is throwing a fault, but it might be purely informational and not super addressable. Or it could be that, you know, there’s some sort of subcritical performance issue that’s going on, but it’s not actually throwing a fault. Um, but it is causing an observable loss.

In your system performance. The second issue could be redundancy. So if comms are lost for an entire site, you could get, you know, a hundred different inverters that are throwing the exact same fault, and then you’re just drowning in alerts because you’ve got a hundred different. Faults for what ends up being like one issue to actually fix.

So on Wattches side, you know, what we’ve done is try to eliminate that redundancy as much as we can through our own algorithms, but then also give you a lot of customizability so that you can actually see what alerts matter to you and determine that at the account level, at the site level, and then also at the device level.

Um, so that really cuts. Through a lot of the noise and allows you to, to really see what’s addressable. And then on the KPIs or performance metrics side, what we’ve seen as a lot of the pitfalls have mostly been using KPIs during, you know, the feasibility study phase or using the same KPIs that you use to secure financing to, for operations and maintenance.

Mm-hmm. And for kind of like ongoing issue identification, um, and. We think that’s not necessarily like the best use of those KPIs. A couple that come to mind are like be be your performance versus predicted. Essentially using your PV cyst to then identify ongoing issues during operation or performance ratio.

We’ve seen that kind of used a lot and the main issues with those are that it can be sometimes hard to determine, you know, what’s a weather related loss versus what’s a system loss. And then within that, which system losses are actually just inherent to the design of the system versus which are actually actionable and require corrective action.

So the way that Wattch likes to approach this is that we really push the use energy Performance index or EPI, otherwise known as performance versus expected. And basically what this is, is just using. A combination of real-time weather data and then a full digital twin of your site that allows you to identify losses not only at the site level, but then at the device level, at the string level, and then also see, you know, what type of loss is it, and actually be able to know remotely.

What the issue is, where it is, why it’s occurring, what corrective action to take, and sometimes even like, is it worth fixing? Um, depending on the amounts of financial loss and what that corrective action costs.

Sean Swentek: You mentioned the term digital twin. Can you explain for the listeners in simple terms what that is and how it really transforms?

Uh, solar monitoring for commercial properties?

Alex Snedeker: Of course. So. I think maybe the best way to approach that question is to first go over what a digital twin is not. You know, we, we often hear a lot of misconceptions about this, that a digital twin is a comparison between one inverter and another inverter on site, or that it’s.

A comparison between past output and present output, or between predicted and actual output. It’s not any of these things. Importantly as well, it’s not ai. So we already know how silicon cells behave. It’s one of the most well researched things, and so we actually use math and physics to come up with, you know, a completely deterministic output.

And so that means, you know, when you see that output, we know exactly like. How the performance model arrived at it. So what a digital twin is, is a combination of three different types of inputs. Uh, the first is a comprehensive model of all of your system components. So basically this is the meters inverters and the strings that are on site.

So we take, you know, all of the models, all of the orientations for your strings, all of your string lengths. How they’re all connected together along with real-time meteorological data, so namely temperature and radiance. And then finally, component performance data. So that could either come from component data sheets or it could come from a third party, such as the CEC.

And so, like I said, we kind of put these through a series of very complicated mathematical equations and then arrive at the outputs. Of expected power, um, but then also expected voltage and current. And again, that’s, you know, at the inverter level and at the string level. So the reason why that’s super exciting is because a lot of legacy monitoring tools are using linear or a top down model.

You only need really a few details about the site, but then there’s a lot less that they can tell you because they don’t have the expected voltage in current first. You know, with a digital twin, the margin of error is a lot smaller. And so you don’t necessarily need like a big haircut or an adjustment to solve for all of the unknowns.

We think that. You know, linear models can be really good for some of the larger sites. Like if you’ve got a ground mount that’s got all of the same inverter models on site, all of the strings are the same link. There’s one orientation. Um, a linear model could work really well for you, but then if you’ve got a rooftop site, which you know, I feel like is gonna be pretty prominent.

Among a lot of CRE developers, you’ve got, you know, multiple orientations on a pitched roof. You might have multiple string links, you might have different capacities of inverter in the same site. Each addition of heterogeneity is going to introduce. You know, a lot larger margin of error unless you’ve got a full scale digital twin that is taking all of those things into account.

For those that you know, really wanna nerd out about this, we’ve actually got a blog post coming out on exactly how we build our digital twins and performance models for, you know, everything that I can’t detail here.

Sean Swentek: Well, I for one, look forward to reading that, so I’ll keep an eye out for it. The other term you mentioned was this energy performance index or EPI.

How does that differ from traditional performance metrics and why do you consider a superior wage or managed solar performance?

Alex Snedeker: Yeah, that’s a great question. So I think first it’s worth defining. What is a good KPI, um, for Solar Performance management, and we see that as being three different things.

First, we want it to be able to be determined in like close to real time, so definitely less than a day. You don’t wanna have to take a month’s worth of data in order to produce this metric. The first would be, you know, it separates. Addressable losses from kind of the inherent or system losses that are designed in.

Uh, and then the third would be it’s easy to tell if a site is healthy. So a hundred percent means that you’re doing fine. And anything less than a hundred percent means there might be something that’s addressable. And EPI has all of these things. So if you’ve got a really good digital twin, then.

Generally, you know, 99% to a hundred percent means that your epi is healthy. Um, and anything less than that kind of answers the question of like. Am I leaving value on the table? How much is that? And then if you add in some of your PPA data through our financial models, we can then calculate in dollars down to a 1% margin of error.

Like exactly how much value are you leaving on the table. Um, so if you wanted to manage by exception, you could do this by, you know, sorting your sites by epi. But you know, a better way might even be to sort it by like how. How great are these losses on a financial basis rather than just a kilowatt hour basis?

And that’s what we’ve actually seen a lot of our customers do as well, is just, you know, come into the tool and immediately sort all of their sites by epi and just go down the line.

Sean Swentek: Can you put that into real world numbers? Like what sort of production value losses are you seeing from renewable energy customers when they’re using, you know, outdated performance metrics?

Alex Snedeker: It’s definitely hard to prove a counterfactual of like, you know, what, what our customers have done if they had not used, Wattch what I can say actually when you’re, when you’re talking about leaving value on the table, it actually reminds me of the study that Omni Ian did around, you know, managing CNI scale pv and the fact that.

It was, I think six truck rolls per megawatt per year was the average that these sites required. And when you think about it, you know, each truck roll could be thousands of dollars depending on how remote that site is and what the remedy is. So. If you can eliminate even one of those truck rolls per megawatt per year, you’ve already paid for your monitoring platform, like right there.

Okay. And we’ve also found that among sites that didn’t necessarily start out with Wattch during the commissioning process, but that might not have had monitoring or might have moved to us from another monitoring platform that we see about an 8% or so improvement on average in their performance. And I think this makes sense when you think about like the increased visibility and diagnostic tools that you’ve got through the platform,

Sean Swentek: talking about variances with like weather related issue versus real performance issues, how can asset owners interpret those performance variances?

And, and how does Wattch support that work?

Alex Snedeker: Yeah, so I think, you know, the first thing that I would say if you wanna diagnose whether it’s weather related or not, is to make sure that you have a source of weather data on site. Um, generally we would recommend having some form of onsite weather sensors that measure temperature and a radiance.

We’ve also recommended satellite data for customers for whom that’s not really feasible. Both of them can work and can support the digital twin, but you have to have some form of. Onsite weather data. I think the second would be making sure that you’re using KPIs that actually take that weather data into account.

Mm-hmm. Um, and so the ones that I had mentioned previously, it’s they might take only a radiance and not temperature, or not necessarily take any form of weather data into account.

Sean Swentek: For those listeners, managing portfolios of assets across multiple sites, what are the biggest challenges these folks face when they’re trying to manage energy across a portfolio?

Alex Snedeker: From what I’ve seen, honestly, overwhelm in a couple of different ways, right? So it. Seems like to date, you know, O&M teams have had to use different tools for each type of asset that they have on site, whether that’s solar or batteries or load monitoring. They have to use different interfaces to monitor and then control those devices, and there’s no real way.

To see and report on mm-hmm. Portfolio as a whole without, you know, pulling out a bunch of data into Excel and doing hours worth of, of kind of massaging it into what you need. What I’ve also seen is that teams are drowning in too much data and they don’t have a really good way of being able to cut through the noise and mm-hmm.

Take insights out of that data to, to help them. You know, do the corrective action that they need. Really, it seems like, you know, these teams just wanna be able to manage by exception, rather than pouring through spreadsheets and having to actively look at every site. And that’s what we are trying to help them do, right, is to allow them to be a little bit more.

Reactive. And in fact, that’s one reason why we don’t have a north star metric of daily active users. ’cause I feel like if you are going into the monitoring platform every day and feel like you have to look at every site, like that’s not good. We want you to, to trust that, you know, when we alert you that something is wrong.

Yeah.

Alex Snedeker: That you have a high degree of certainty, you know, around the fact that that actually is an issue. What I’ve also seen is that there are a lot of. Workforce constraints as well. So it’s, it’s hard to hire and train someone into some of these positions. Um, and then of course, you know, deployment is accelerating, so you’ve got a small O&M team that has to take care of more and more projects.

So really, most of the, the O&M teams that we are talking to, it’s like the same persona of. They’re so overworked, they’re so stressed, and so I think there’s really a hunger for just better data and, and better insights.

Sean Swentek: How do you quantify financial impact of under performance to help owners prioritize, you know, which issues should I address first?

Alex Snedeker: Yeah, so in Wattch, at least you can do. Using your PPA rates and set up PPAs and apply them to either one site or a portfolio of sites. And then from there you can kind of see how that shakes out in a couple of different ways on our site page. Like I was saying before, you. See the financial impact of any underperformance.

And so in addition to the like A through F Health score that we have and epi, we also have a financial health score where you can see that dollar value and be able to rank it among all of your sites. Um, we also have that. Same metric on the alerts page so that you can see, you know, what is this alert costing me every day?

And then be able to prioritize your alerts that way. And then that’s really great because you can use that to determine if you’re. Corrective action is actually gonna be beneficial or not, or what the hour OI might be, because you could see an alert and decide like, actually it doesn’t make sense to address that.

Like given the cost of addressing it and the, and what it’s costing me per day.

Sean Swentek: Now, how can property owners use your monitoring technology to accelerate deployment of new projects?

Alex Snedeker: I think in a couple of different ways. One thing that we haven’t necessarily talked about yet is via faster commissioning.

So you know, when I was at Asset Management North America, I looked, or I listened to a panel on challenges in the CNI space, and I remember Forefront, which is one of the panelists mentioned. Commissioning as being super important for de-risking the performance of their assets.

Mm-hmm. Um,

Alex Snedeker: why? Because it’s the last chance that you’ve got to, to have your installer on site before that final payment.

And I know coming from CNI development that after that it can be very impossible to get people back out for corrective action. Um, so what’s really cool about having a good digital twin is that then you can see. See in minutes and not like in days or in weeks if everything is installed correctly, and then hopefully fix any sort of issues that you have the same day.

And by contrast, we’ve actually seen assets come online with. Mistakes that we think happened during installation and they were never caught. Um, and then it just becomes a lingering issue and they don’t really notice it until the, the asset switches into Wattch. And then by that point, it’s not even worth the cost of corrected action.

So now you’ve got a site whose performance is handicapped indefinitely. We’re able to catch mistakes in the commissioning process. So, you know. First, it’s like helping people commission faster. Um, helping people catch mistakes on the front end and then, you know, making it easier to catch mistakes on an ongoing basis means that you can lower your cost corrective action.

You can. You know, optimize your performance and then hopefully you’ve got more capital to deploy on other projects and, and you’re deploying these projects way faster. So hopefully, you know, the right monitoring platform can be a catalyst for being able to, to grow your deployment.

Sean Swentek: You mentioned Solar Plaza’s am an event that a lot of us were at recently in San Diego.

I’ve been asking everyone I talked to after what, for what was, what’s your take on sort of the top trends? What are on people’s minds that are managing assets out there? What? What was the talk of the show in, in your mind?

Alex Snedeker: Yeah, that’s a great question. I, you know, being a monitoring provider, I think I only frequented those panels that were really around like addressing O&M issues and pv and so yeah, I remember that story kind of catching me.

There was another one as well about being able to diagnose issues and. Have and, and like apply the appropriate corrective action to them. Um, and so everyone literally, I think literally every panelist was saying, you need a digital twin. And so that was incredibly validating to us because it can both, you know, cause people to overspend in some areas and underspend in others if they, you know, aren’t able to separate out those losses.

So yeah, I was very much heartened by that and really I think. You know, the, the role of technology, like you were saying mm-hmm. Very much growing in terms of being able to put together a tech stack that can help accelerate deployments. And, and it was funny because a lot of panelists said like, you know, there’s.

There’s no platform necessarily that can do it all. Um, and to a certain extent I agree, but my hope is that, you know, we’re continuing to add on more use cases in the O&M and asset management space that can kind of make, Wattch a more comprehensive tool along the lines of what a lot of these people are looking for

Sean Swentek: as asset owners are evaluating their tech stack.

Do you have like three questions every renewable asset owner should be asking of their like, current monitoring provider around performance metrics? Like, what are the things they should be growing them on to make sure they’re, they’re getting what they need?

Alex Snedeker: You know, I think making sure to have some sort of performance versus expected is an absolute must.

And most monitoring platforms that you’ll encounter today already have some version of epi performance versus expected, so I’m not too worried about that. Questions that I might ask to follow on would be, first, how do you calculate it? And then digital twin would be the answer that I would hope for on that.

And you know, the target, I think should be having a one to 2% margin of error. If you’ve got onsite sensors with that full digital twin, I might then ask, you know, what are the inputs to your digital twin? And. The, the answer that I might look for there is, you know, the electrical layout, having some source of realtime I radiance and temperature.

Mm-hmm. Um, and then having that third party performance data to underlie it. And then the third question that I might ask is, what are the outputs from? Because again, great to be able to get expected power at a site level, but I think you should be able to also get expected voltage and current being able to get it on a per inverter and per string basis, um, which just really enables you to be able to diagnose issues remotely, to be able to kind of pinpoint which pieces of hardware are the issue and then.

Not only avoid diagnostic truck rolls, but then when you actually do have to roll a truck, you’re showing up with the right people, the right tools, and you know exactly where to go on site. So you’re not looking around for an hour trying to figure out, you know, which piece of hardware is the problem.

Like I said before, like monitoring should be able to super easily pay for itself, but I think that can only happen if it delivers good insights and it’s able to to show you what might be happening on your site remotely.

Sean Swentek: Completely agree. Alex, anything else I didn’t ask you that you were hoping to be able to speak to about Wattch or the industry at large today?

Alex Snedeker: Really, I was gonna say, I think we, we went over all the, the questions pretty thoroughly.

Sean Swentek: Amazing. If people wanna learn more about you and Wattch, and the work that you’re doing, where’s the best place for them to go?

Alex Snedeker: Yeah, they can always reach out to me at [email protected]. That’s S-N-E-D-E-K-E-R at WATTCH.io.

You can always go to wattch.io to learn more about some of our capabilities. That’s also where you can find the Wattch insights. Blog and you can connect with us on LinkedIn as well.

Sean Swentek: Excellent. Alex Snedeker from Wattch. Thank you again for joining me today on the show to all my listeners, thank you for tuning in to another episode of Future of CRE Sustainability.

I’ll see you next time.

Alex Snedeker: Thanks again, Sean.