Experiment: Are US Utility Scale Plants In Danger of Not Meeting Financial Contracts?

Experiment: Are US Utility Scale Plants In Danger of Not Meeting Financial Contracts?

By: Adam Baker

Case Study in Freelance String Monitoring Shows Utility-Scale Solar Sites Don't Have the Correct Info to Identify Site Problems

How do you determine if the output from your utility-scale solar plant meets your financial pro forma? If it's underperforming, do you feel your monitoring system gives you the information you need to understand why?

I would argue that 95% of utility-scale solar sites in the U.S. do not have the correct information to identify problems on their site. And it’s costing them.


The Stigma of String Monitoring in the U.S.

The clear answer to the problem is a detailed DC health check-up. But string monitoring (as it has traditionally been available) has a bad rap in the U.S. It’s a belief held by most owners that string monitoring increases the cost of a plant, and generates a LOT of data, but the return on investment is hard to pin down, especially because it’s not easy to look at a large set of data and compare what’s different today from last week, or last month, or last year.

Then again, this perception is only partially true.

Yes, a ‘permanent string monitoring solution’ is a very expensive capital expense, either when installed during construction or retrofitted to a site after construction. Let’s do a quick calculation with $10 per analog data value as an estimated cost for integration into SCADA. 24 inputs per combiner, with 10 combiners per MW, equals $12,000 that must be added to a budget. That’s probably an overall 20% increase in instrumentation cost.


But what if solar owners simply paid for a temporary string monitoring solution to analyze their data a few times each year?


Temporary String Monitoring Significantly Reduces Costs

Until now, a temporary string monitoring solution hasn’t been available, so solar plant owners have compensated with other, less precise site monitoring solutions.

  • Zone Monitoring: Most U.S. PV sites use zone monitoring, which looks at the input to the inverter from each combiner box. If the site is rectangular with identical arrays on flat land, you might be lucky and have an expectation of 200A at each inverter input. Needless to say, I haven’t seen too many of those sites. The Southwest’s wetlands and limits of disturbance create irregularly shaped site layouts, and developers end up adding a few strings at one combiner, and taking a few away at the next. If one combiner should have 240A, and the next should have 170A, how would you know if underperformance is occurring? (See also: The Importance of Data Normalization). In short, zone data will tell you when something in a general area is very wrong, but won’t provide any details.
  • SCADA: Most plants believe their SCADA system is the end-all-be-all of monitoring. While SCADA is an excellent monitoring tool, temporary string monitoring goes a few steps beyond. SCADA doesn’t outline problems, normalize data, or plot an ideal power curve throughout the day that goes beyond clipping. String monitoring can.
  • Commissioning: Others rely on commissioning to identify site problems. The problem is, commissioning agents only take a snapshot of what was happening just before the plant was turned over to operations. With a one-time reading of current at some moment when irradiance was > 500W/M^2, It’s highly improbable they’ll have the time or detail to trigger a detailed visual inspection of modules on a 10MW site. Commissioning ends up being a pass/fail analysis on if the site is working.


How Temporary String Monitoring Works

A temporary string monitoring solution works in much the same way as permanent solutions, but with a small operating expense rather than a larger capital expense.

Here’s how Affinity Energy’s Solar String Analysis system works:

  1. First, we use clamp-on current transformers (CT) to record current data from each DC string feeding a combiner for a short period of time. Clamp-on CTs are extremely quick to install, and require no tools or site shutdown. Ideally the data would be taken from one clear sky day, but the same results can be derived with a few partly/mostly cloudy days as well.
  2. Then, CTs are moved to new strings every few days by our engineers, or by site O&M to collect new data from different parts of the site.
  3. Lastly, Affinity Energy analyzes the collected data with specific metrics to identify the low hanging fruit that’s making the site underperform (e.g., blown fuses, lose or melted MC4 connectors, small shading, dirt accumulation, cracked cells).

The solution is non-contact and low hazard. There are no SCADA requirements, so whatever is already installed will work. Analysis is performed offline, so you don’t need a full-time operator to provide analysis.


Case Study: Temporary String Monitoring

We decided to put our prototype solution to the test by testing our temporary string monitoring equipment with the DC strings from our own building’s rooftop solar setup. Could a temporary string monitoring solution be a viable option to improve PV reliability and O&M activities by enhancing the data needed to make critical decisions about maintenance?

Who: Affinity Energy’s 30kw rooftop installation in Charlotte, North Carolina


  • 3 PowerOne inverters
  • 4 DC strings
  • 11 - 260W modules per string
  • ~1.1 DC/AC ratio
  • Inverter 1 max output: 7,280W
  • Inverter 2 max output:~9,700W
  • Inverter 3 max output: 9,900W


  • Find the cause behind the 2,620W difference between inverter 1 and 3, and the 200kW difference between inverter 1 and 2
  • Determine if temporary string monitoring could find failures and ultimately assist in increased energy output


  • Install a CT on each DC string
  • Collect new data from the CTs every 5 minutes, with over 100 data points per inverter, per day
  • Normalize and analyze data to calculate each string’s impact to energy
  • Determine likely causes based on performance differences


Case Study Findings

The biggest glaring error we found with temporary string monitoring was that an entire string was not operational. After isolating and examining the affected string, we located one melted MC4 connector.

By looking at historical data for our rooftop’s solar output, we could see the impact to the total kW output of the system. After analyzing our Duke Energy bill, we determined that melted connector cost us $60/month for 18 months.

It was also news to us that two of three inverters were affected by tree shading on the edge of the lot, and that the very small (<2 degree) east-west pitch of the roof biased a detectable amount more energy to the west, from 2 inverters in the afternoon.

In summary, the repairs we implemented from the data found by Solar String Analysis helped increase the energy yield of Affinity Energy’s rooftop by 8%.

I wouldn’t expect this outcome at most utility-scale solar sites, but it’s a pretty good start for our prototype.


Beyond Rooftops: The Cost of Underperforming Strings to Utility-Scale

Let’s scale the problem we found on our rooftop to a real solar site.

A 1 MW site is 33 times bigger than our 30kW rooftop. If we multiplied the $60 lost monthly by 33, that totals $1,980.

As a commercial installation, we pay $.14/kwh. Assuming the owner of a utility-scale solar plant pays about $.07/kwh, that small, one-string melted MC4 connector would cost $1,000/month/MW.

Over the 20-year lifecycle of the plant, that totals almost $240,000/MW.


Why Should You Care About Temporary String Monitoring?

There’s a lot of strings at a 5MW solar site. How could a typical O&M tech find one bad string among 1,200? Plus, if the site is still making full power, what’s the loss?

There are probably an average of 6 MC4 connectors crimped on jumpers and harnesses between the strings and the combiner. Each one of those is an opportunity for failure. That means 7,200 crimps at a 5MW site must be performed without a failure.

If your electricians perform at 4-sigma, there will be at least one bad MC4 connector on your site. Human performance at 4 sigma is already challenging, but when you factor in the outdoor environment (e.g., weather, freeze-thaw cycles) I would expect to find 2 or 3 marginal harnesses per MW if the site is doing well.

Overall, bin class variation can explain up to a 2% variation from one string to the next, but anything bigger than 6% is as if you lost an entire module in each string. For every percentage lost, you narrow your clipping window. Fixing small problems will increase DC current, which will broaden the shoulders of your clipping window, which increases energy…Energy that you were not being paid for. Energy you could benefit from every day of every month of every year.

If I could get a 100x return on my investment, I’d be comfortable investing a few bucks. Wouldn’t you?

Let me know if you’re interested in talking about string monitoring for your site.

Adam Baker - Affinity EnergyAdam Baker is Senior Sales Executive at Affinity Energy with responsibility for providing subject matter expertise in utility-scale solar plant controls, instrumentation, and data acquisition. With 23 years of experience in automation and control, Adam’s previous companies include Rockwell Automation (Allen-Bradley), First Solar, DEPCOM Power, and GE Fanuc Automation.

Adam was instrumental in the development and deployment of three of the largest PV solar power plants in the United States, including 550 MW Topaz Solar in California, 290 MW Agua Caliente Solar in Arizona, and 550 MW Desert Sunlight in the Mojave Desert.

After a 6-year stint in controls design and architecture for the PV solar market, Adam joined Affinity Energy in 2016 and returned to sales leadership, where he has spent most of his career. Adam has a B.S. in Electrical Engineering from the University of Massachusetts, and has been active in environmental and good food movements for several years.