Expert
Solar PV
Design

Why Your Solar System Never Hits the Numbers

A client signs off on a 131 kWp system. The yield forecast looks great. The install goes ahead. Then the system goes live and it never hits the numbers.

Here's why.

We modelled this commercial rooftop twice in PVSol. The first design ignored the chimneys on the roof entirely. The numbers looked clean: 1,004 kWh/kWp, 98.3% PR, 0% shading loss.

The second design modelled them accurately. The real picture: 835 kWh/kWp, 81.8% PR, 9.8% shading loss.

That's a gap of 22,184 kWh every single year quietly missing from the client's return year after year for the life of the system.

The client didn't get a bad system. They got a poor design.

This is what happens when shading objects aren't modelled properly. The forecast is inflated, the ROI calculation is wrong and the client spends years wondering why their system underperforms.

Accurate 3D simulation — actually putting the chimneys, rooflights, plant and parapets in the model — isn't extra work. It's the job.

If you want a design that tells the truth about what your system will actually generate get in touch.

enquiries@colwallenergy.co.uk

The Grid Said No and Nobody Checked First

A developer lines up a 250kW rooftop. The design is sound, the client is keen, the price is agreed. Then the DNO connection offer lands and the grid will only accept 94kW.

Suddenly a signed off project is a third of the size, the numbers don't work and someone has to go back to the client with bad news.

This happens more than it should, and it is almost always avoidable.

The local network only has so much room. If the area is already busy with generation, or the fault level headroom is tight, the DNO will cap what you can export or turn the application down altogether. None of that shows up on the roof. It shows up on the network.

The fix is simple. Check the capacity before you design, not after. The DNO publishes heat maps and network data that tell you what a site is likely to accept. Size the application to that reality and you stop promising a client something the grid was never going to give.

I check capacity upfront, size the G99 to what the network will actually take and handle every conversation with the DNO through to the connection offer.

If you want to know what a site will really connect before you commit to a design get in touch.

enquiries@colwallenergy.co.uk

The Bigger Battery That Made Less Money

A site owner asks for the biggest battery that fits the budget. Bigger has to be better, right?

Not always. We modelled the same site with two batteries. A 100kWh unit paid back in a little over five years. A 215kWh unit on the same site took closer to nine.

The extra capacity sat half empty most days. It cost more to buy and rarely got used, so every extra kWh dragged the payback out instead of pulling it in.

A battery has two numbers that matter. Power in kW is how fast it can charge and discharge. Energy in kWh is how much it holds. Get the balance wrong and you either can't move energy quickly enough or you're paying for storage you never fill.

The right size comes from the site's actual half hourly demand and the tariff it sits on, not a rule of thumb or a round number that sounds impressive on a quote.

I size storage from real consumption data, match the power and the energy to how the site actually uses electricity and show you an honest payback before you spend anything.

If you want a battery sized to make money rather than fill a spec sheet get in touch.

enquiries@colwallenergy.co.uk

Half Your Solar Is Going Over the Fence

Here's a quiet problem with a lot of commercial systems. Nearly half of what they generate is sold back to the grid for pennies while the building keeps buying power at full price.

A system sized on roof space alone will often make far more at midday than the site can use. The excess gets exported at a few pence a unit. Meanwhile every evening the same site buys electricity back at four or five times that.

That is not a system fault. It is a sizing decision made without looking at the demand.

The money in solar is self-consumption, using what you generate on site to avoid buying it. Every unit you use yourself is worth the full import price. Every unit you export is worth a fraction of it.

Model the array against the site's half hourly demand and you can see the split before anything is installed. From there you either size the system to the load or add storage to soak up the middle of the day and release it later.

I model PV against real demand, size for self-consumption and show you exactly how much of your generation you would keep versus give away.

If you want a system built around what your site actually uses get in touch.

enquiries@colwallenergy.co.uk

Where Your Yield Actually Goes

Most solar quotes get the yield number wrong. Not by a bit. Sometimes by 15 or 20 percent. And the client only finds out two years later when the system underperforms and nobody can explain why.

Yield is the number everything hangs off. Payback, ROI, the business case you put in front of a finance director. Get it wrong and the whole thing wobbles.

Here's what actually moves it.

Orientation and pitch. South at 35 degrees is the textbook answer but almost no commercial roof gives you that. East west split arrays often make more sense on a flat roof because you spread generation across the day and match it to when the building is actually using power. That matters more than chasing a peak number.

Shading. One dormer, one flue, one neighbouring building at the wrong time of day and a string drops off a cliff. This is where a proper simulation earns its money. PVsyst models the near shading hour by hour across the year. A spreadsheet estimate cannot.

Soiling, temperature and losses. Panels lose output as they heat up. They lose more when they are dirty. Cable runs, inverter efficiency, DNO constraints all take a slice. Add them up honestly and the gap between the brochure figure and the real figure is significant.

Self consumption. High yield is worthless if you export most of it at a poor rate. The best design matches generation to the site load, and that means looking at the actual half hourly demand data, not a guess.

The number you want is a P90. That is the yield the system beats 9 years out of 10. Design to that and the business case holds up under scrutiny. Design to a best case P50 and you are hoping.

We build every design around real data and honest losses. No inflated headline figures to win the job, because the client remembers who quoted the number that never showed up.

If you want a yield assessment that stands up when the system is running, get in touch.

enquiries@colwallenergy.co.uk

The PR Test That Tells You Nothing

A lot of people in the industry can't tell you what a PR test is. That's a problem.

Performance Ratio is the standard measure of how efficiently a PV system converts available irradiance into usable energy. Test it at commissioning and you have a baseline. Skip it and you're handing the client a black box that will cause issues moving forward.

Having said this, we know standard PR has a well-known weakness in that it doesn't account for temperature. A system tested on a cold clear day will post a flattering PR figure that tells you nothing useful about summer performance. That's where Power AC PR, or PAC PR, comes in.

PAC PR normalises output against temperature-corrected expected power, giving you a figure that holds up regardless of when the test is carried out. It's a far more honest measure of whether a system is actually performing as designed, and it's the metric we use.

But none of it means anything if the design was wrong to begin with.

Contractual PR targets can be baked into the modelling from day one, so there are no surprises at handover.

Good design. PAC PR at commissioning. A documented baseline. It's not a high bar. It's just something a lot of people don't understand, and it should be mandatory at commissioning.

If you're an installer who wants to get this right, Colwall Energy can help with design support, PAC PR testing or both. Drop us a message at:

enquiries@colwallenergy.co.uk

What Do You Want the Battery to Do?

"Can you add a battery to the design?"

We get asked this on a daily basis. Our first question back is always the same.

What do you want it to do?

Usually there's a pause.

A battery sized for self consumption looks nothing like one sized for peak shaving. Add grid services or tariff arbitrage into the mix and the numbers change again. Get the use case wrong and you've either spent too much on capacity that sits idle or bought a battery too small to do the job it was actually needed for.

Before anyone talks kWh we need certain data so the battery can pretty much size itself.

If you're specifying PV and storage and want a design that starts with the why rather than the what, get in touch.

enquiries@colwallenergy.co.uk