Green Building Forum - Solar tracker with MPPT data

Solar tracker with MPPT data

GBP-Keith — Fri, 02 May 2008 18:46:58 +0100

The first (almost full month) of data is now available for the Green Building Press's 750 watt array on a solar tracker. I'm really excited about how things are working out. Just take a look at the results below bearing in mind it is still only April. I hope to help prove that the anti-PV lobby are wrong and it is one of the most significant technologies available to us.

First of all. Here is the tracker that the data is collected from.

There are six 125w BP polycrystalline solar panels all wired up as a single group at 72volts but with a variable peak of up to about 150volts. this gets converted back to battery voltage (nominal 24 volts) by the MPPT controller just before being added to the bank.

Solar tracker with MPPT data

GBP-Keith — Fri, 02 May 2008 18:57:56 +0100

Here are the first results (April 2008). The yellow plot in the top chart (just for comparison) is the 480wp static array that I've been monitoring separately. I have included this for comparison.

The first three flat days were prior to installation.

The middle chart shows the output of both arrays in kWh/day.

However, the bottom chart is normalised from both arrays at ah/m2 and I think is shows quite clearly that the tracker is generally putting out a fairly consistent 40-50% improvement over the static array and more on the really brightest days.

Solar tracker with MPPT data

GBP-Keith — Fri, 02 May 2008 19:00:13 +0100

Here is an extract from part of the next update of Green building Bible 4th edition that shows payback times are falling fast because of rising cost of grid electricity. the sums are simple enough and tracker technology combined with MPPT would probably (based on my early monitoring) improve this perhaps by 30-40% more:

"The installed capital cost of a PV array is approximately Â£400m2 for commercial scale and Â£600m2 for domestic buildings (the lower cost of commercial systems arises due to economies of scale). 8m2 of PV cells would give an approximate peak output of 1kW and when using a normal charge controller could produce approximately 900kWh of electricity each year at an installed cost of Â£3,200 and Â£4,800 respectively (depending on installation method and solar access). On the other hand, if the same amount of electricity was purchased from the grid at 15 pence per kilowatt hour then the cost would be Â£135 per year offering a financial payback of 23 and 35.5 years respectively. Of course we are currently witnessing regular large rises in the unit cost of electricity from the grid. If this pattern continues then the payback equation will continue to improve in favour of PVâ€™s. "

Solar tracker with MPPT data

GBP-Keith — Fri, 02 May 2008 19:05:44 +0100

The outback MPPT controller cost Â£425.00 which is about twice the cost of a normal controller suitable for this size array. However MPPT offers savings in cable sizing OR enables installation farther away from the batteries or control system and obviously gives better output, thought to be about 25% better. I'll post a picture shortly showing the way in which the collection and delivery of the energy is controlled.

Solar tracker with MPPT data

GBP-Keith — Fri, 02 May 2008 19:26:16 +0100

Here are some install costs on this particular kit.

MPPT Controller = Â£425
Our particular tracker is hard to cost because I bought is second-hand but the new cost would be roughly Â£1,200 + delivery cost to UK
The wattsun prices are here: http://www.wattsun.com/prices/Wattsun_Tracker_Prices.pdf

Six BP solar 125 watt panels (6.1 m2)= Â£2,800
Installer = between >Â£800< (this will vary on how much work you can do yourself)
Pole and concrete = Â£200
Cable terminals and fuses - Â£600

Total roughly = Â£6025

That comes out at about Â£980m2

After 1 year of monitoring we will be able to see how it compares against a static array for payback.

Solar tracker with MPPT data

stephendv — Sun, 04 May 2008 16:38:48 +0100

Interesting. It would be helpful if the data between the tracker and the static array was normalised so that it was a like for like comparison. Have you considered adding an MPPT to the static array? This would also make it clearer whether the best gains are from the solar tracker or from the MPPT charger.

Solar tracker with MPPT data

GBP-Keith — Mon, 05 May 2008 19:24:19 +0100

Thanks Stephen. I' have now added a comparison chart that compares the two arrays per sq metre (see above). On your other point, I have plans to add an MPPT to the static array but to confuse it even more I have two static arrays and both will be put on the same controller.

Solar tracker with MPPT data

GBP-Keith — Wed, 07 May 2008 09:38:58 +0100

Am I misreading this or does the comparison of ah/m2 actually show a 100% (daily) gain by the tracker over the static array not the 40-50% I thought. I had better double-check my data!

Also below that is the same data but kWh/kWp as requested by StephenDV. As far as I can see, no matter how you cook it it comes out the same - amazing.

By the way, the two arrays are identical technology - polycrystalline. The only difference is the age. The static array is about 15 years old so I think I'll get a test put on it.

Solar tracker with MPPT data

stephendv — Wed, 07 May 2008 15:40:35 +0100

Gawd damn! :) That's a spectacular difference. On the sunny days (22/4) the tracker + MPPT does even more than 100%. You've made me rethink my off-grid design now.... Any chance you have the same data for the autumn/winter months? I'm definitely sold on the MPPT, the value for money proposition is just too good - but I've been hesitant about a mechanical tracker as I thought it wouldn't be as effective since the position of the sun doesn't change that much in winter - which is when we really need the power.

(What's going on on the 11th and 12th? Looks like the values don't correlate...)

Solar tracker with MPPT data

GBP-Keith — Wed, 07 May 2008 17:57:15 +0100

Well spotted Stephen. There will be occasional blips and usually on un-sunny days when it gets windy as I shut off the tracking control and anchor the tracker with it's face angled into the wind (if that makes sense) which may make it a less perfectly angled collector than static on those occasional days (although it still did 50% more, this is probably showing the MPPT alone at work). The most obvious example is probably the 19th. I will double-check the data but I assume this was what was happening. I think that from now on I will not the tie-down days and duration.

Unfortunately I will not have any autumn/winter data until the end of the year as I am brand new to this monitoring lark. I will prepare a kwh/m2 chart shortly.

Solar tracker with MPPT data

stephendv — Thu, 08 May 2008 09:07:00 +0100

There's a discussion about your data on the navitron forums: http://www.navitron.org.uk/forum/index.php/topic,3300.0.html
A comment that came up was that kwh/m2 might not be an accurate comparison if the PV panels aren't identical. Maybe a kWh/kWp (total rated W of the panels) would be better?

Solar tracker with MPPT data

GBP-Keith — Thu, 08 May 2008 19:02:54 +0100

Chart added above as requested Stephen. It tells just about the same story.

Solar tracker with MPPT data

GBP-Keith — Sun, 01 Jun 2008 19:09:17 +0100

May was a little disappointing sun wise but here are the results. Again the static array is included for comparison.

Solar tracker with MPPT data

GBP-Keith — Tue, 08 Jul 2008 17:20:27 +0100

June data with the static array shown for comparison. The start of June was nice and sunny with the end being better for wind energy, unlike June 2007 which barely saw any breezes all month.

Solar tracker with MPPT data

GBP-Keith — Tue, 08 Jul 2008 17:22:13 +0100

And June data ah/m2

Solar tracker with MPPT data

Rachel — Tue, 08 Jul 2008 19:50:27 +0100

oo Keith! I want one. The solar panels I have on my roof are not south facing, but south west. In the mid winter the sun only reaches them for an hour or two... such a waste for an off grid system.

Solar tracker with MPPT data

GBP-Keith — Tue, 05 Aug 2008 16:13:33 +0100

July 2008 data with the static array shown for comparison. Luckily there was a moderate amount of wind again.

Solar tracker with MPPT data

GBP-Keith — Tue, 05 Aug 2008 16:15:24 +0100

And July data ah/m2

Solar tracker with MPPT data

GBP-Keith — Thu, 02 Oct 2008 19:39:03 +0100

August data. Disappointing sunshine (almost non-existant) this month removed some of the advantages of the tracker over a static system.

Solar tracker with MPPT data

GBP-Keith — Tue, 07 Oct 2008 19:21:00 +0100

September data. It is clear to see form this months data that when the sun did come out the tracker with the MPPT got stuck into collection, whereas, regardless of the sunnier days later in the month the static array failed to really take advantage of the situation.

Solar tracker with MPPT data

JackyR — Wed, 17 Dec 2008 20:33:47 +0000

Keith, did you ever add that MPPT to the static panel? Also, what happened when you tested it for ageing?

If these figs really are normalised for the above, I'm going to forget about static panels (in the UK) and jump straight to the tracker. Terrific stuff!

Solar tracker with MPPT data

Ed Davies — Thu, 18 Dec 2008 12:07:07 +0000

I'd be very interested to see the data for the most recent months. I suspect that over the winter the difference between the tracker and the static PVs will be much less, though the MPPT might distort things here.

Solar tracker with MPPT data

stephendv — Thu, 18 Dec 2008 14:41:55 +0000

JackyR, last time I compared the cost of a tracker with the cost of additional panels on a static array - the additional panels won. You can compare output via: http://sunbird.jrc.it/pvgis/apps/pvest.php?europe=

If you're off-grid then that changes things too - since the tracker would be less effective in winter, but the additional static panels more so. And in an off-grid system where you have a constant consumption throughout the year, it's the winter months that really determine the size of the system.

Solar tracker with MPPT data

PaulT — Fri, 19 Dec 2008 13:02:29 +0000

As a link to building regulations:

SAP assumes a 1KWPeak array can provide 800Kwh pa (no shading and location dependent)

A 750Wpeak array, South facing, 30 degree and no shading has an allowable 625 Kwh provision per year (Wales)

Interestingly the emissions displaced are 0.568 Kg CO2/kwh - This is a much more realistic value than the 0.422 Kg CO2/kwh used elsewhere in SAP and does give some interesting results:

If you use 1 kwh of electricity and generate 1 Kwh of electricity you get a credit of 0.146 Kg CO2!

This means that there is a bias in SAP for using PV as apposed to using less electricity (just an observation)

A major issue is that SAP assumes all types of pv behave the same - that ther is a fixed relationship between the KwPeak ratign and the annualised output.

In the context of "zero carbon homes" this has serious consequences - a developer requires only a high Peak ratign and does not need to consider other performance factors; Clearly a lot more research needs to be fed into SAP to rectify this.

Solar tracker with MPPT data

Ed Davies — Fri, 19 Dec 2008 14:03:53 +0000

Posted By: PaulTIf you use 1 kwh of electricity and generate 1 Kwh of electricity you get a credit of 0.146 Kg CO2!

This means that there is a bias in SAP for using PV as apposed to using less electricity (just an observation)

That might be a feature, not a bug. If the PV tends to generate electricity at times of day when high emission power stations (coal) are running but domestic use is typically at times when the grid is running mostly on gas, nuclear and a bit of wind then that could plausibly reflect reality.

A major issue is that SAP assumes all types of pv behave the same - that ther is a fixed relationship between the KwPeak ratign and the annualised output.

Do you have any data on how much difference the different types of PV make?

Solar tracker with MPPT data

PaulT — Fri, 19 Dec 2008 16:46:17 +0000

I agree that it is a feature thather than a bug.

I do not have data on different PV types - I studies them at CAT around 4 years ago with Rob Gwillam (Eco engineering pioneer and hero to generations of students), appologies for forgeting detail:

I do know that there are different responses to light levels so that power output vs light intensity is not always a linear graph - it may be that lower peak output pv could produce more accumulative energy in lower light conditions.
Researchers into PV are just as interested in this as the peak output and a lot of Nano based research is all about getting as few protons as possible to move electrons across the substrates.

Another factor is mathcing of individual panels - they all respond differently - in affect an internal resistance/ responsiveness - a power output curve . A good supplier / manufacturer willl match individual segments of an array either in parallel or in series so that they respond in the same way...

Can get very deep on these types of issues, but quite right considering the cost... Roll on cheap Nano technology with consistent power curves.

Solar tracker with MPPT data

GBP-Keith — Sun, 21 Dec 2008 11:51:04 +0000

Here is the October data form the tracker and the static array. I have normalised the vertical axis to match the best available output for the year. I will adjust previous charts to all show the same vertical axis when I have time. I should point out (as i have before) that the tracker spends all windy days (over 17mph) in non-tracking mode and anchored.

Solar tracker with MPPT data

GBP-Keith — Sun, 21 Dec 2008 12:53:48 +0000

November 2008 data. Though they now look much closer on the charts, (due to normalisation of the vertical axis), the tracker (on average) is still doing almost three times that of the static.

MPPT or tracking? I think a combination of both but it is clear that the tracker can get to the sun when it is there and I think this his the significant tip to the average over the month.

Drawbacks to the tracker:
Risk of damage in high wind - I strap it down when it gets windy. Not everyone would be happy with going out at midnight with a torch in mid-winter but i quite enjoy the opportunity to get involved with the system in such ways.

Newer models are probably much more robust.

Solar tracker with MPPT data

Ed Davies — Sun, 21 Dec 2008 13:47:19 +0000

Thanks Keith, this is really useful data.

Here's my take on your data normalized to hours of nominal panel output per day. In most cases I took the your average Ah for the month, multiplied by 25 volts (guestimated battery charge voltage) then divided by the panel nominal output (480 or 750). For the May tracker/MPPT data point you didn't give the Ah number so I averaged the daily kWh figures and divided that by 750. This seems to give a suspiciously out-out-of-line result.

Month	Static	Tracker/MPPT	Gain
March	1.65
April	2.2	4.08	1.85
May	2.41	5	2.07
June	2.77	5.06	1.83
July	2.43	4.17	1.72
August	1.72	2.82	1.64
Sept	1.73	3.07	1.77
Oct	1.29	2.15	1.67
Nov	0.69	1.07	1.55
Average	-	-	1.76

Gain is the multiplication of nominal hours which the tracker/MPPT combination gives.

Solar tracker with MPPT data

hotelRefurber — Sun, 21 Dec 2008 14:18:12 +0000

Keith, Some questions:

1. can you please summarise as I am slightly confused, how many kWh's have you generated on average per month so far, for each type of system, the number of sq m and the instal cost?
2. do you have maintenance or depreciation on your batteries? I know you have told me they were from a submarine, but not everyone has access to that kind of stuff...