I agree with that beautifully coloured set of plots @Simon1D 

The power-curve for an array varies all the time according to density of cloud cover, temperature, birds sitting on the panel and (worst of all) leaving poop behind!

For this reason the inverter will reassess the power-in and adjust the current being drawn (eg re-plotting the curve according to a new data set). After all, it has no idea that I’ve connected a 320W panel. I might’ve just thrown a breaker and added another pair of panels in series!

From observations I’ve made on commercial inverters, this process of re-checking the input power seems to occur in intervals of up to 6 seconds. Between this interval, the current being drawn remains constant even though the panel may be capable of outputting more power (eg cloud cover lifts).

This might be a good moment to show a diagram which illustrates the amount of energy available from the sun and how this translates into what is achievable using current silicon panel technology.

The effect of cloud cover is quite pronounced. Any shading on an array of PV Panels will dramatically reduce the power available.

Agreement is always good :-)

As well as reading the wikipedia page, I also dipped into the literature. Well, I scanned one student’s review article …

This is already fairly big business, and growing fast, so I must assume that proper assessments have been done as well, though they’d inevitably be commercially confidential. I can’t resist sharing yet another idea for a MPPT algorithm which comes to mind - I have no doubt that someone has already tried it, but I didn’t see it mentioned - perhaps @Transparent might have come across a description somewhere.

With apologies for going all academic, I’ll outline the idea here rather than in a PM...

How easy is it to vary the load the inverter presents to the PV panel?

If the load is more or less arbitrary, then I’d suggest superimposing a small sine wave component on a constant background value (the load would be the sum of these two).

This small ripple in the load would cause matching ripples in the panel voltage (and current, and power). All the frequencies are known, and some signal processing can extract their amplitudes and from there the local slope of the P(V) curve is easy to calculate. This in turn can be the basis of an MPPT algorithm, with the constant part of the load being adjusted to increase/decrease the panel voltage V depending whether that slope is positive/negative.

Sounds simple enough. I’ll find it hard to resist setting up a little model to play with...

If you fancy having a go at making a model MPPT unit @Simon1D then there are several online projects, mostly based on an original design by Tim Nolan.

One that was described in detail is here on Instructables. and uses an Arduino microcontroller to charge a 12v lead-acid battery from PV panel with a 25v output. But note in the introduction that the author met a significant problem in 2016 and ceased development.

It has subsequently been taken up by others with mixed success, either because they understand electronics but write poor code, or vice versa.

These projects should not be undertaken lightly. The hardware and software needs to fail safe when there is a fault. There is a significant amount of energy being stored which can do a lot of damage if you don’t get the battery charging regime correct!

Ah no - sorry to disappoint, and just to correct any misunderstanding - I’m talking about a software (or mathematical) model.

Well that’s a start I suppose. Next we’d need someone to turn it into code for an Arduino and someone else who knows which is the hot end of a soldering iron

To be honest, the thing that disappointed me most in what I read about MPPT was people’s apparent willingness to go ahead with this or that idea as the basis for an MPPT with minimal regard for whether the underlying algorithm was well founded.

But at least we’ve moved on from the significance (or otherwise) of where the current and power curves intersect.

 

 

Wow, the amount of energy we get some PV panels suddenly seems pretty small.... What about this Solar Edge method which @EverythingNeedsAUserName said was superior?

What potential is there to improve this with technological improvements?

@Tim_OVO i think you'll find that I said SolarEdge claim superiority, not that they are superior ;) 

Well I’m the one who said that Solar Edge was superior in this post 

I think we now have just enough information here to consider why this could be so.

Solar Edge divides the functions of the classic string-inverter into two separate components.

Each PV Panel (on the roof) is fitted with its own Optimiser. This unit uses an MPPT algorithm and operates as a DC-DC converter.

The outputs from all the optimisers are connected together as a ‘string’ and feed into the inverter box within the house. The Inverter is left only with the task of changing the DC input to 230v AC, synchronised with 50Hz from the Distribution Grid.

Each Panel can therefore have different electrical characteristics. Moreover, one Panel in shadow doesn’t reduce the output of the entire string.

Depending on the amount of shading or East-West alignment which the rooftop arrays are subjected to, a Solar Edge system can increase the total output by anywhere between 2-30%.

To answer @Tim_OVO’s preceding question, if the rooftop arrays are in full bright sunshine then the Solar Edge technology will provide minimal advantage. But if the arrays are partially or fully shaded at times of the day, then the average output will always be better than using a classic string inverter.

Installing a Solar Edge system might cost around 10% more, depending on what proportion of the work is taken up with scaffolding and access issues. The main drawback is that you can’t mix & match with any other manufacturer. Either your system and its monitoring is fully Solar Edge, or it’s not.

@Transparent thanks. That makes sense and helps me understand why it is sensible for our East/West arrangement. The west-facing panels aren't as bad in the morning and the east-facing aren't so poor in the evening. Good job we had sensible installers. 

@Simon1D wrote:

To be honest, the thing that disappointed me most in what I read about MPPT was people’s apparent willingness to go ahead with this or that idea as the basis for an MPPT with minimal regard for whether the underlying algorithm was well founded.

If you could inspect the proprietary algorithms within the Solar Edge system, I think you would be much relieved. The entire system seems to have been based on good R&D work by a team of innovators. It wreaks of high-quality engineering.

This is just the sort of project which UK designers should now be looking to emulate in our quest for technologies to combat Global Warming.

It’s now 20 years since there were two major breakthroughs announced in the field of Solar Panel design:

• Synthetic porphyrins, replicating the photosynthesis model; research jointly by Osaka City University and University of Sydney
• Thin-film printed electrodes; research by several USA universities into panels created by a process akin to an inkjet printer

Neither of these looks likely to threaten the dominance of silicon panels any time soon despite lab experiments having reported energy yields of around 40%.

The current focus of academic research into harvesting sunlight seems to be for the production of hydrogen. That’s useful, but can only ever be contemplated on an industrial scale. There’s no safe way for domestic production to pressurise and store hydrogen!

Sounds good. I tried to read between the lines of what SolarEdge do say publicly, and felt encouraged.

I think I can see how such a system works:  each PV module not only has its own MPPT (to optimise power output) but also has its own DC-DC converter. Internally, the modules in a string may be passing different currents, because their characteristics or levels of illumination differ) but externally (on the other side of the converter) the currents are all the same and so they can be connected in series.

As far as I understand it (again), that current passing through one string is more or less arbitrary and I assume that if multiple strings are connected in parallel, as per your earlier graphic, then those currents are adjusted to make the voltages the same for all the strings (and at a level that suits the inverter).

I hope that this general principle remains open for exploitation by other companies, and isn’t somehow restricted to SolarEdge because of some IP protection. (I’m not an IP expert either.)

With all this DC-DC conversion already going on, I’m surprised that any battery charging doesn’t happen directly, instead of going via the inverter and back again (90% round trip efficiency is still only 90%, after all...)

This surprises me too - feel like I’m finally starting to get all this AC/DC business (just watched the film ‘The Current War’ which I’d recommend for anyone wanting a beginners introduction to the historical competition between Alternating Current and Direct Current). 

@Transparent any beginner’s friendly explanation as to why you’d need a DC-DC converter?

I had solar panels installed in 2011, I have 12 Yingli 235W panels which feeds a Fronius IG 30 Inverter with a 20yr extended warranty, it is cabled down the void my soil pipe runs in to the garage to the Elster generation meter and from there to a 16A MCB in my Consumer Unit.  The total peak power rating is 2.8kWp.  To date, I have generated 26.58MWh.

When installed I was getting 41.3p per kWh generated, index linked for 25 years, this year I’m getting 56.03p.  The break-even point was some time in February 2020 so everything from then is profit.  My current export rate was 3.95p per kWh until I joined the V2G trial with Electric Nation and now I get 10p per kWh between the peak hours of 16:00 to 19:00 and 5.5p per kWh at other times.

I have an Optimmersion intelligent immersion heater controller which was installed in July 2014 and after 7 years has paid off 89% of its cost of £300 so will probably take another 7+ yrs to break-even; this is calculated by estimating gas savings from not having to heat water with the boiler.  However, there’s a certain satisfaction to getting free hot water and in the summer we turn off the gas boiler and use the hot water the immersion provides.

Following the acquisition of my first Nissan Leaf in April 2018 I had a zappi EV charger installed in June 2018 for the princely sum of just under £80 after the government grant and hard won contribution from Nissan.  The zappi automatically charged the car when surplus generation equalled or exceeded 1.4kW and was a great device.

@PeterR1947 thanks for your figures - I’m still working out if solar panels are going to be worth the cost. We have a Pod Point EV charger which only works at 7kWh so unless I get 20 panels (which I won’t!) the charger would draw twice as much power as the panels give. Funnily enough the solar panel supplier said they don’t recommend changing the EV charger to a Zappi, even though they could supply it !?

Is there an easy way to work out the return on investment (ROI) time horizon? @juliamc will not be alone in wanting to know if the cost is worth it based on what is no doubt a long list of factors.

Asking the installer might leave you doubting the answer. Most people are too busy to do the research and calculations needed to work it out themselves.

I’m imagining some self service online tool where you can specify the size of the roof, direction it faces, location, price range and a nice round figure is generated…… wishful thinking? @sylm_2000 @EverythingNeedsAUserName @PeterR1947 

On day one, I created a spreadsheet and recorded the generation every day with the aim of moving to weekly or even monthly…..I’m still doing it daily!

I then compared the total cost of the system (£12710.31 inc. extended warranty on inverter) with the total income generated.  I have a daily record of generation since 2011 compared to the estimated generation, also monthly and annual generation compared to estimated generation and also the weather on the day, graphs of annual generation compared with the original estimate, graphs showing all years cumulative generation, table of monthly generation and payback, a table of FIT payments.

I feed my data into a website https://pvoutput.org which allows you to choose favourites and compare your output against theirs and I have to say mine holds up very well against some larger systems.

That's what The Energy Savings Trust’s Solar Energy Calculator sort of does. It uses Post code, Roof slope, direction, shading, installation size, but you've got to adjust for actual or quoted  cost.

There's no simple calculation. At least, I've not found anything despite extensive research. There is a STANDARD from the government within the Standard Assessment Procedure( SAP 2012) which is for Energy Efficiency in building, and also includes EPCs. 

It is astoundingly complex, 

The electricity produced by the PV module in kWh/year is 0.8 * kWp * S * Zpv where all those are further detailed. For example, S 

There are some interesting stats on the government website. 

Solar photovoltaics deployment

Monthly deployment of all solar photovoltaic capacity in the United Kingdom.

New figures monthly. July will be released on 26 August. 

During June 2021, there were 5,234 installations, another monthly record figure since the closure of FiTs in March 2019. The majority (83%) of these new installations were sub-4kW installations. 

Solar photovoltaic (PV) cost data

This table contains information on the cost per kW of solar PV installed by month.

The average cost of solar installations in 2020/21 was slightly higher than in 2019/2020. For the smallest (0-4kW) installations, the mean cost increased to £1,628 per kW installed. 

 

It is astoundingly complex, 

The electricity produced by the PV module in kWh/year is 0.8 * kWp * S * Zpv where all those are further detailed. For example, S 

 

You weren’t wrong about the astounding complexity, @EverythingNeedsAUserName! - I wouldn’t have considered that the slope of the roof can also have an impact.

Out of interest, do Solar panel suppliers/installers often give a rough ROI timeframe? And if so how accurate have these given timeframes turned out to be in reality? 

Great to hear you’re keeping track of how the estimates are comparing to the what’s actually been generated, @PeterR1947 - what’s the general trend - better or worse than expected?

I’m guessing it’s always worth doing independent calculations if possible but sure I’m not the only one who would get slightly thrown off by the maths!

The slope of the roof is almost important as its orientation towards south @Jess_OVO 

Here’s a diagram from an excellent online tutorial published by Viridian Solar in Cambridgeshire

My own roof faces directly south at a pitch of 35° 

The LPA in my area advised that the same pitch be adopted on an estate of 750 houses being built in the locality. However, they are unable to enforce that because developers need not comply with local requirements on matters of energy nDeregulation Act 2015, section 43]. Only National planning policy requirements for energy generation and conservation are mandatory.

The developers have therefore used 45° pitch roofs because this requires less roofing material.

HMG are unlikely to legislate on this issue because they regard mass construction of houses as a primary route to grow the UK economy post-Covid.

MPs don’t tend to appreciate the practicalities of what’s required to combat Climate Change.

The slope of the roof is almost important as its orientation towards south

Here’s a diagram from an excellent online tutorial published by Viridian Solar in Cambridgeshire

 

Great reply this, @Transparent - another possible silly question: why not adjust the tilt of the panel? If the roof is not between this sweet spot of 25-50 degrees, surely a tilt of the panel on the roof can adjust for this?

@Jess_OVO as part of the sales process we were given a calculation of the estimated ROI (Return On Investment) and Payback Time. The estimate at the time of installation was an 11 year payback*, but at this rate it will be 8 years. 

I imagine that something similar is done under the current scheme. 

*I see this document says 8 years 11 months, so perhaps I misread it.

@Tim_OVO We were told that there is more risk of damage by weather (to the panels and to your roof), nesting birds and so on if the panels are not close to the roof.

We had a roofer fixing an unrelated gulley and he commented on the closeness of our panels and on the lack of bird mess /nests! He says he often sees pigeon nests on the top of and underneath panels. 

16 to 20 years with the smart export guarantee? Does this sound in the right range now with all the caveats around site and usage? 

https://www.greenmatch.co.uk/blog/2014/06/solar-panels-are-they-worth-it

22 to 60 years without the SEG?

Should we really think about domestic retrofit solar pv now as a way of limiting climate change rather than having a ROI?

Interesting if this is correct. Makes we wonder if it is better from a climate change and ROI basis to invest in up coming green companies rather than retrofit domestic solar panels? A somewhat controversial question i admit.

Or why not invest the same amount of money in a solar farm or wind turbine array?