The Dos and Don’ts of Data Monitoring

In this 20-minute webcast, we discuss:

• What is the latest in data monitoring technology and how does IoT come into play, or does it?
• Why is testing important and what’s in it for me?
• Why do I want all this data?
• What should I consider when choosing how to network my solar system?

Watch the recorded webcast:

Webcast Transcript:

Hello everyone and welcome to Solar Tea Time!

I’m Craig Noxon, VP of Enterprise Sales at REC Solar and we have our REC Solar Manager of Monitoring, Craig Merrilees, as our guest expert this month. Welcome Craig!

Thanks namesake, It’s my pleasure to be on.

It is good to have another Craig on the call, I truly appreciate you being here.

Before we get started, a little housekeeping. I wanted you to all be aware that your line is muted but you can type questions inside of the webcast under the questions tab. We will try to answer questions at the end of the webcast or we will get back to you via email. OK, that’s it for the logistics. Let’s get to the topic of the day, Solar Data Monitoring!

Understanding solar monitoring is crucially important for anyone managing a solar system. It can save you a lot of money and a lot of grief. So, without further ado, let’s share the splendors of data monitoring with our audience. First question:

What is Solar Data monitoring?

At its most basic level, it’s all about gathering data from various devices at a solar installation and saving that information to a database in the cloud. We collect data from revenue grade meters, the inverters, and we gather information on the amount of our sun’s energy either from on-site sensors or satellite observations. In the database we typically record all this information in 5-minute intervals.

So first step is to gather the data, I get it. Before we go further how do we do that?

We have industrial computers that we install onsite called data loggers that query the devices and send that info to the cloud. It’s all part of the monitoring package that gets installed along with the inverters, meters, and modules. REC Solar partners with a renewable energy monitoring company that provides us the hardware and the use of their cloud based database and web servers. The end user, either in the form of an REC Solar service associate or an REC Solar customer, gets a login and can access the 5-minute data. There are reports that can be setup to send automatically on production, and alarms and alerts that are configured to tell us if something goes awry.

And what if the data logger Internet connection is down?

Good point, well actually in that case, the data logger is designed to cache data locally. It continues to read data from all the inverters and meters. Then it will send that data up to the cloud when the Internet connection is up again.

How long can the data logger keep info without the need to send it to the cloud?

Ah computer memory is plentiful these days. We can easily store a month or more of data on the data logger before it’s needed to be sent on its way. And in a worst-case scenario, if the connection can’t be fixed in an acceptable amount of time, we can get the data via the old school sneaker net method. You know, when you walk up to the computer in your sneakers and get the data off it manually. Then we would upload it to the cloud.

Now that we have all this data in the cloud – what do we do with it and how can it help a customer?

The data is the “tell-all” for knowing if the system is producing as it should. We know for example, that given a certain amount of the sun’s energy falling on the array, a certain amount of energy should be flowing out of the inverters. It’s called a performance test and it works a bit like this. At solar noon it’s typical that the sun will provide us 1,000 watts per square meter of power, reaching the Earths’ surface. This is on a clear blue-sky day mind you. A module itself might be rated for 300Watts at 25 degrees Celsius. Let’s say you had 3,000 of those modules making up an installation with a DC capacity of 900,000 watts or 900Kilowatts. Well… There’s one more piece we need before we can actually predict what energy will flow from the system. We need to take into account any losses that occur between the modules and the point where the electricity gets placed in to the facility. There are losses in the DC conductors, the inverters, and AC conductors. Typically, this represents a number of around 10%… and in monitoring circles is referred to as the de-rate. To continue with our example, if we had that 1,000 Watts per square meter of sun energy falling on our 900 kiloWatt solar system at 25°cell temperature, then we would expect the 10% decrease of the 900kW DC to results in 810kiloWatts of power being fed into the facility. If the sun is lower in the sky providing us with less power, say only 500 watts per square meter, then we’d adjust down our prediction by 1/2.

There are a variety of ways this performance test can give insights into the health of the solar installation. If we score this by taking the actual energy over the expected energy we can get a performance index. 100% being ideal. If this number gradually goes down over the course of a couple of weeks or months, it’s a good indication that we have a soiling issue. Speaking of soiling, looking at the performance index score over time, if you see an uptick in performance around the time of a big rainstorm, then you know your modules where cleaned for you by mother nature. It’s also true that if you get a small amount of rain that washes the dust down to the bottom portion of the module it can actually make production worse. You’d see that in the data too.

Yes, I know clean modules play an important role in the performance of an installation. And now I understand that monitoring can help me decide when to clean. On to our next question.

What is the latest in data monitoring technology and how does IoT come into play?

Wow Craig, IoT, Internet of Things. Busting out with these tech-head acronyms on me are you? Well let’s see. You could consider the Inverters at a solar installation like a smart appliance you might have at home. But instead of your dryer letting you know when your clothes are dry or your refrigerator telling you it’s time to buy more milk for Solar tea time… Your inverter tells you if there’s an issue with your solar energy generation system. The inverters can detect utility grid disturbances, and /or report on faults that might arise.

Great, so that means I get to maximize my investment in my solar by knowing if it is producing as it should be.

You got it. There’s another side to this Craig. You see, there are some compliances that need to be adhered to that are mandated by the various utilities. It’s outlined in UL 1741. It started off with a safety feature that when utility power was removed from a solar installation, that that site would stop producing electricity. The idea here is if utility workers are working on the transmission lines, we want to make sure there is no foreign source of electricity on the lines from Solar generation facilities that could pose a danger to the workers.

How would an inverter know if the utility grid went away? I mean if there are many inverters all on site producing electricity, how can the inverter tell the difference from neighboring inverter’s electricity vs the utility electricity.

Ah, great catch there Craig. There are a variety of ways that this is accomplished, but here’s just one as an example… it’s called frequency pushing. In North America the electricity that we get from the utilities is AC, meaning alternating current at 60 hertz. Essentially what this means is the voltage of your electricity is moving from positive to negative 60 times per second.

Yeah, I get we’re 60 hertz and Europe uses 50 hertz, so how does that help the utility worker?

Well, imaging if all the inverters were trying to make the frequency just a little bit faster, always. But you have this big national utility grid out there that is not movable. All the gas and coal and nuclear generators spinning at a set frequency resisting the change. All of the inverters are trying to push this thing faster. Now imagining if the grid goes away, all of the sudden the inverters DO push the frequency faster than 60 hertz. Now the rules of UL 1741 come in to play. They state the inverters can only run within a certain frequency range. The inverters exceed that tolerance and shut down.

What does this have to do with data monitoring?

It’s more than just data monitoring. It’s about the network. The whole IOT thing. These inverters can get new programming remotely now. It’s like Tesla pushing out updates to their cars with over-the-air programming. This is huge, in the past we would’ve sent a field tech to the site to visit each and every inverter, connect a laptop and upload an update , you know, to change the programming for an inverter., but now with the advent of IOT, they can be reprogrammed remotely. So as the compliance rules change, so can the programming of the inverters. It’s the network that is at the core of this ability. Take Hawaii for an example. With the huge percentage of Solar compared to the rest of the generating sources, the implementation of UL 1741 supplement A is being required. Supplement A adds all kinds of beneficial rules, this time not just to protect the utility worker if the grid goes away, but now the inverters can SUPPORT the grid during times of brownouts and sags. We are creating a more reliable grid by adding renewable generation. This inverter magic can now happen thanks to the reprogramming made possible by the same IOT technology that allows you to turn your lights on at home from your smart phone while watching your dog play at the dog park.

I get it. The data monitoring and the inverter remote updates have a common theme of the underlying network.
OK, well that’s a good transition to this question…

What should I consider when choosing how to network my solar system?

Well… you need to know what level of reliability and performance you require at the solar installation. If we’re talking about just reporting meter and inverter data up to the cloud, then a caching data logger with a cell modem is sufficient. But… if we’re talking a site where the utility requires the remote control of the inverters. And let’s say those inverters need to respond to the control signals within a second or two? Then we’ll need to consider a dedicated land based Internet connection with possible redundant fiber optic rings at the site. It really depends on the need at hand.

There are basically two types of networks that can exists at an installation. There’s the tried and true RS-485 industrial serial network which is slow but reliable and great for long distances. Perfect for getting data from an inverter a thousand feet away. Just make sure to use the proper RS-485 cable and use terminating resistors at the ends of long runs to prevent signal reflection. RS-485 devices need to be wired in a daisy chain fashion with a maximum length of 4000 feet, but never in a star configuration. Then we have the more modern TCP IP network which is what the Internet works on. It does get wired in a star configuration and with Ethernet switches can go 300 feet in length, or with fiber can go miles. We also utilize wireless point to point and mesh networks when it makes sense to.

Here’s another question for you, Why is testing important and what’s in it for me?

Testing is core for us here at REC. We do monitoring testing during initial site commissioning to verify the network performance, and to check the quality of the data from the various devices. We make sure the data from the revenue grade meter… matches what the inverters say. If we have onsite weather sensors, we make sure they’re calibrated, and match hand held instruments we use during the commissioning process. We gather data via the monitoring system and score the solar installation based on a few days of production. Once we get approval from the utility and we’re confident that the system is producing as expected, we Cristen the site as ready for production. Once the PV system is up and running, the energy data as recorded in the revenue grade meter, at times, can be used for SREC credits and rebates with certain utilities.

We also want to make sure any shading has been properly dealt with if any exists on-site. During the testing stage we look at the array to see any shading that occurs during the day. It’s important for the shade to be limited to as few strings as possible. For optimized production, we’d rather see 50% of one string be shaded rather than 25 % of two strings. It’s just how the modules react to shade, a little bit of shade has a big negative impact on the production of the string.

But isn’t it true that systems designed with optimizers or micro-inverters are immune to this string shading issue?

Yes, that is absolutely the case. And for shade prone installations, REC uses optimizers and microinverters for maximizing site output. If you look at monitoring data in play-back mode for a microinverter installation you can pretty much see clouds go over the array as the different modules get shaded and produce less. It’s a pretty cool sight. Speaking of clouds, weather is one of the largest contributing factors to Solar performance. Some years just have more solar irradiance than others and it’s a very localized phenomenon. We use visible spectrum Satellite data which has been enhanced with infrared imagery to improve the accuracy of the Satellite data. It has to do with the satellite being able to tell the difference between say snow and clouds. Visibly hard to do, but in the infrared spectrum, doable.

How does the Satellite data help the customer?

We use the Satellite data coupled with the monthly production model to see if the site is producing compared to the original engineers’ modeled production. It’s just another benchmark we use to see how a site is doing. Basically, we know what a site should produce for a month given certain predicted weather conditions. We adjust this expected amount by the difference in predicted weather vs actual weather as seen from the Satellite. Again, a score of 100% is ideal. Another method we use that is quite powerful is the inverter heat map analysis. This works great on sites with multiple inverters or optimizers where you can compare the output of all to each. The “heat map” color coding provides an effective visual way to see the high and low performers.

Why do I want all this data?

It all boils down to generating the most energy. To do this we need to identify losses. If it’s soiling, shading, inclement weather, or equipment issues, we want to understand what impacts our installations. By analyzing monitoring data, and doing various performance tests and inverter comparisons, we can maximize the energy by learning if there are production shortfalls, and what is responsible for them. That… and it makes for a great activity while sipping a glass of solar tea.

Thank you, Craig, for sharing your years of insights on how customer can leverage data monitoring to their benefit. We are out of time for this month’s Solar Tea Time. If anyone would like to talk to Craig, feel free to email us. You can also email us if you have questions that you would like answered in future Solar Tea Times.

On behalf of everyone at REC Solar, thank you for joining Solar Tea Time and happy holidays!
Cheers!