How can we account for PV degradation and maintenance?

Back in the heady days of high Feed-in-Tariffs (FiT), when the UK had a Government backed payment scheme for promoting PV panel installations, I took advantage and invested in my own solar array.

Fifteen years on, and the panels were caked in grime and algae. Generation from the panels was close to half. I knew they were in a bad state when I had an email from the FiT payment company, asking if my system was OK!

For a bit of context, I’m based in a very rural location. There’s not an abundance of trees in my part of the world, but there are plenty of fields, lots of agriculture and a decent amount of lakes.

I wouldn’t expect an urban PV array to experience such a high level of algae. Most panels are self-cleaning for day-to-day dirt, but this stuff was stuck on like glue.

The state of my roof before work began to thoroughly clean it. Courtesy: Cotswold Roof Cleaning

So, to get me into the spring cleaning spirit, I’ve had the panels professionally cleaned by Cotswold Roof Cleaning.

The photos speak for themselves, and I’m looking forward to seeing my generation figures jump up in the coming month.

What a difference a day makes! The same shot of the same roof, just six hours later.

But this raises a thought about the way we model onsite generation in SAP calculations and whether we can use HEM to improve accuracy.

When we create a SAP model of a house that has PV panels, we enter four key bits of information:

• the size of the array (measured in kilowatt peak (kWp)),

• the orientation (to the nearest 45 degrees),

• the tilt of the roof (to the nearest 15 degrees), and

• an overshadowing factor, ranging from 1.0 ‘my roof is free from shadows’ to 0.2 ‘why are you bothering to put panels here?’

SAP calculates the amount of energy that will be generated (based on average weather data from Nottingham – regardless of where your site actually is).

And then, SAP cuts 20% off the total.

It assumes that all PV arrays operate at 80% efficiency due to dirt, even from Year 1. The SAP assessor has no way of over-riding this.

In my case, thinking back to Years 1-5, my PV array should have been running at near 100%.

Fast forward to Year 15, it was running close to 50%. But is that a normal level of degradation?

Apparently not. The hive mind on LinkedIn shared research papers from the USA that show arrays only lose 0.5-1.0% of their capacity a year.. so it sounds like my array is an outlier, and not in a good way!

The Home Energy Model, which will be replacing SAP from next year for Building Regulations compliance and EPC generation, gives assessor far more control over the technology that a house may use to create and store energy.

The current version of HEM (which isn’t the finalized version so is subject to change), asks the following:

• kWp – same as SAP

• Orientation – to the precise degree

• Tilt of roof – to the precise degree

• The size of the array and how far it is above the ground (not sure of the relevance of this)

• Rear ventilation (whether air can flow behind the panels are not – panels are less efficient the hotter they get).

• And other buildings that may cast shadows can be recorded, improving the accuracy of the overshadowing factor.

These inputs are an improvement over SAP, but the impact of dirt and degradation still isn’t being considered.

It’s not yet known if HEM will apply a standard 20% 'dirt factor' as SAP does, or whether the assessor will have greater control over this.

We also have to consider that not all arrays are cared for equally... With the new Approved Document Part L targets expected to push for PV on all new homes, how many of these arrays will be suitably maintained in years to come, and how will HEM assessors be expected to consider this?

Interestingly, when you enter a battery into HEM, it asks how old the battery is, and assumes a degradation factor based on the age.

If the model can consider the drop-off in performance of an old battery, surely we can apply the same logic to a PV array?

Granted, few homes will suffer algae blooms to the extend I've had to deal with, but maybe a 1% annual reduction is a good place to start?