A different problem to be solved: Solar Power Technologies President and CEO Ray Burgess on SPT's Clarity suite and power optimization for large PV plants

Solar Power Technologies President and CEO Ray Burgess; Image source: Solar Power Technologies
Solar Power Technologies President and CEO Ray Burgess; Image source: Solar Power Technologies

Ray Burgess joined the Solar Power Technologies team as President and CEO in July 2009. He has over 30 years of leadership experience in the technology industry, spanning semiconductors, software and micro-mechanical systems. Prior experience includes TeraVicta Technologies, Tao Group, Freescale Semiconductor, Motorola and Texas Instruments.

Solar Server: Let's start with an overview of your Clarity suite of products, and what they specifically offer to your clients. What are the main barriers to system performance that your Clarity products address?

Ray Burgess: I will start by referring to something in Solar Server's April 2011 piece on microinverters and power optimizers . Where we came from two years ago when we started the company, our objective was to do an optimizer product for larger-scale systems. As we got into that and got down to development for prospective clients, it became clear that there was not a lot of credibility in the claims of energy harvest that would come in larger scale systems for an optimizer or microinverter technology.

The headline claims of 20-25% gain just aren't there. They are not there unless you've got badly compromised or shaded arrays, or arrays that are misaligned, and that just doesn't happen in larger-scale arrays. So we got into this believing that there is a different problem to be solved.

With Clarity, what we are trying to do is optimize harvest, maximize return, and minimize the risk for people that own, operate, finance and insure large-scale arrays. And that's a very different type of topic. It ends up not being specifically a hardware problem; it ends up being as much a data management problem.

So we do need best-in-class hardware. We have an optimizer, a DC optimizer, and we also have a standalone panel monitoring device.

But more importantly, our value for large scale systems comes from our scalable mesh network, site server, and our cloud-based computing -it’s really a data management system that would kind of create intelligent arrays.

We make arrays intelligent, such that they adjust themselves where they can, but tell the owners exactly what, where and when to intervene to maximize performance, based on business rules the owners set.

I liken it to the dashboard lights that you have in a car that tell you that your oil's too low, your water bottle is empty, your brake pads need replacing. Basically you have to have someone who set the basic sets of rules, and then the system tells you when something needs to get done.

We're going to get to the point here where what we've introduced in Clarity is a system where large-scale arrays can adjust and optimize themselves, if optimization technology is deployed., Then you have a cloud-based computing environment that is telling the O&M companies how to be the most efficient they can in keeping those arrays at optimal performance.

The Clarity system does all of that. It's not just a hardware product, it's not just a monitoring product and it's not just an optimizer product, though those two things play a part.

Our focus is on large-scale systems, and when I say large-scale systems, really, economically we are talking about 100 kW up to the largest arrays that have built in the world, several hundred MW.

We focus there because we are thinking of people where output is key, financing is key, , and they've probably got contractual commits in terms of PPAs, or minimum output guarantees from the O&M companies or the owners.

So performance matters, thresholds matter, and increasing output by one or two percent can well improve the financial performance of an overall project by some 10% or more.

If you look at what these companies face today, the people that own, operate, finance and do the O&M for large-scale arrays, have no meaningful technology options.

All of the optimizers and microinverters don't scale. You said that in your April article, and  that's quite right. Microinverters get prohibitively expensive once we get above a few tens of kW. And all DC optimizer technologies today are limited by the way that they handle data, and the way they gather information, and the fact that they have to be controlled from a central point.

And most of the monitoring systems that are out there today have issues in terms of handling and presenting data. First off all they are plus or minus five percent accurate, which means that you can't achieve much in terms of aggressive management and their data management philosophy is all about viewing data, viewing graphs, or listing the output streams.

It requires people to manage that information on an ongoing basis, and large site owners and large O&M companies are frightened to death of data overload.

And so if you think about the options that we have today, they are quite restricted. So that's why we have come up with Clarity.

 

Solar Server: Can you talk more specifically about how Clarity works and what it does?

Ray Burgess: At the base there is panel-level hardware. At the moment, that's an optimizer and a monitor, or ranges of optimizers and monitors.

Our optimizer is a full buck-boost DC optimizer. It's fully autonomous, and it does its own MPPT calculation for the the panel, without any central control. And because it is full buck-boost, you have a lot of design flexibility in a large-scale commercial rooftop for instance, you can do string stretching and stacking, and be completely inverter-independent.

We can handle inverters that do their own MPPT, but at the same time we can operate in fixed-voltage mode and can manage the bus voltage, and we can do this all at very large scale.

The other thing about our optimizers is they can be partially deployed. We think that is critical in large-scale systems, where owners wish to maximize ROI by only optimizing areas of the array with known potential issues It's unlikely that you are going to get payback in a 10 MW array, by putting an optimizer on every panel.

However, in a large rooftop array, you could have 10% of the panels that are potentially compromised by shade at certain times or seasons so you can deploy optimizers there, and no-where else.

By the way, at about 8 ounces in weight, our optimizer is the smallest and lightest in the industry. Our view is ultimately to get deployed inside the J-Box, but at the moment is it a retrofit-able product.

The other panel-level product that we have is a monitor. The optimizer has full monitoring capability within it, but the monitor takes that capability out of our optimizer and packages it in a separate device.

Our products monitor voltage, current and power, to an accuracy of 0.5% 256 times per second. We average those into one-second blocks, and send them through the mesh to the database in the cloud. So we have a very accurate insight into how every panel in the system is working.

This data is all normalized by temperature and irradiance, so we can really see how panels are performing and  if they are going out of warranty - either their manufacturers warranty for the first few years of life, or their power output warranty over 20 years. So we can do panel degradation analysis, but we can also do array soiling or partial-array soiling analysis, so you can implement a threshold-based cleaning strategy. Also note, by having sub-second analysis of every panel, we can also detect wiring faults, arc faults, ground paths and other problems that can occur.

By the way, just like our optimizers, our monitors can be deployed at less than 100% panel level. You can do every panel to get 100% granularity and insight into an array, but you can also do, say, half the panels in the array, you will lose a bit of the granularity, but you will still have most of the features and benefits. In fact, if you have one panel-level monitoring device deployed on a panel in every string, you have the most accurate string-level monitoring system in the industry.

String-level monitoring today with hall effect sensors has an accuracy of about +/-5% across the full range. We are at about ten times more accurate than that, +/-0.5%.

Our hardware strategy, is to have optimizers that can be deployed everywhere, or just on panels or strings that have a problem, monitors that can be deployed on every panel, or all the way down to one per string or on a statistical sample of panels and the ability to mix the two in an array. This gives you  a very flexible deployment capability, based on a cost-benefit analysis that can be performed on a particular site.

Because you can decide what technology to deploy, this gives you a tremendous range of cost options. At the one end, to deploy one monitor per string, and get extremely accurate string monitoring system plus degradation and soiling analysis, you can be down at USD 0.02 per watt. Go all the way up to 100% optimization of an array, and you can be around USD 0.20 per watt.

So we have a range of solutions that scale everywhere in between. Depending on how much optimization, how much monitoring and how much granularity you can scale between those two figures and have a solution that is most optimally designed for the issues that you are going to have on your site.

 

Solar Server: Can you talk about some of the other differentiating features of your Clarity products for O&M companies and site owners?

Ray Burgess: We just discussed the concept of flexibility, the other thing that I will note that is quite different is scalability. We've developed a unique proprietary mesh network, so all of these devices are nodes in a mesh that scales up to the MW scale.

With one small gateway we can talk to an entire MW worth of panels, that's very different than the other solutions out there which probably scale to a couple of hundred panels.

And then of course in the cloud we have this intelligent array management suite. That's where we take all the data from all the panels, and perform extensive analysis and diagnostics based on business rules set by owners and O&M companies- The cost of a truck roll, the cost of a panel, the threshold at which they want to be told to clean the array, and the number of panels that need to be out of spec before we will declare warranty claims.

Our arrays are analyzing all the data constantly against a set of rules, and then alerting the O&M company and the owner the things that need to be done, but only when business rules have been triggered. It then issues a complete maintenance schedule for all the things that need to be done, as well as what the payback is for doing them.

So you don't have to look at charts and graphs and Xcel spreadsheets to work out what is going on. The array is monitoring itself, and then telling you, the owner, or the O&M company what to do.

 

Solar Server: What are SPT's initial target markets, both in terms in terms of geography? Is your company emphasizing the U.S. market, or are you looking at European and Asian markets as well?

Ray Burgess: So we are absolutely going for the global market. Our products have been certified, safety and emissions wise, for all global markets. Initially, j we are primarily operating in North America, that is the U.S., Mexico and Canada, however we have started to deal with a few companies in Europe and North Africa, although to be honest we will probably go through agents as opposed to direct resources there in the short term so we don't over-reach our resources as a small company.

 

Solar Server: Excellent. So it's pretty clear what your advantages are, in terms of the difference between your performance product and your competitors, but what are your strategies to gain market share?

Ray Burgess: There are a few ways we are looking at this.

First of all we've got string monitoring that is ten times more accurate, with extra benefits such as panel degradation and soiling analysis, for significantly less than hall-effect sensing string monitoring today. So there is a pure cost-benefit play, as a way to get traction at this point.

However, we also have a number of companies that we are working with, larger-scale multi-site owning companies, that have arrays where they currently have serious issues. They would love to see if we, because of the optimizer/monitor flexibility, granularity and the intelligent array management capability, if we can determine how we can get these arrays back up to par.

What we are doing initially, with the larger customers, is try-to-buy. We are deploying our system on their sites -so we demonstrate real benefits in existing sites that have issues and if they like it, then of course they buy it and we move on to multiple deployments across their sites.

We can do this because our products are easy to retrofit, on all, or just part of an array.. This allows prospective customers to try the technology so they see the benefit, as opposed to just making claims, as other technology providers do.

We are trying to build credibility by standing behind out product's performance, which is gaining us a lot of support.

If there are three words that I could use to get across what we do, it is flexibility, scalability and intelligence.

Flexibility, because our products can be mixed together and deployed at whatever granularity that you want. With any other solution, you have to do the entire array or it won't work.

The second thing is scalability. There are no other optimization solutions that can scale to large plants, with many thousands of panels that need to be monitored or optimized What we've done with our mesh is scalability with no new wires, which is quite crucial to plant managers.

And intelligence: Having an array tell you what to do, instead of just providing tons of data and graphs. It only tells you what you need to know and when you need to know it. I think that is really different for a large site owner.

These three things summarize what’s different about what we do, compared to every other solution in this field.

Conducted by Solar Server International Correspondent Christian Roselund