Better harvest from the Solar Roof!

There is one benefit of having passed the winter solstice...  bit by bit we shall see improved average harvest from our solar roofs! (although for a couple of weeks either side, the sun's altitude only varies minutely).

My roof is performing better than predicted by the JRC solar calculator, even allowing 10% efficiency losses. The harvest is 7 percent better in Oct, 40% better in Nov and so far, 45% better in December..... (December has been sunnier than average) .
This is pretty good, especially as our winter sunrise is a full hour later than the real sunrise due to our being so close to Sharp Hill. and the morning is our only chance for harvest. Air temperatures are colder than average for December, so has this been enough to improve the performance of the panels?

A day of adjustment - fixing errors

Dec 23: a Horrible day, entirely my fault, although the nice thing is that the sun shone all day and made the work warmer. I wish I had been more precise about measuring the centre of the group of panels, I just worked from the centre-pipe which I thought was central. But it shows yet again how important accuracy is. The left hand array should have been 2 cm to the right, and the right array would have been better if it had been 4cm to the right. In each of the arrays, the pair would have been better a centimetre closer to each other. Seems small, but matters a lot when trying to fix all those bulky 40mm pipe fittings and stiff rubber hosing, with all the jubilee clips required.
I was close to taking the whole lot off and drilling yet more holes in the rails, to refix them.... but thankfully, there was enough 'give' in the fixings to move them sideways, enough to fix the stiff rubber hosing.  By now it was dark, and I hope all those Jubilee clips work well! They certainly make it easier to fix the hoses!
Later in the evening, I fastened aluminium angles to the Polycarbonate front panels (large!) using pan headed screws... this all went well, although its difficult to handle such large panels on my own.

Four Panels up!

21 Dec: Four panels up now. The curvature is because they are dry inside, so the front is expanding thermally, but not the back. Its lucky I have 8mm bolts holding them on in 12mm holes - so there is enough play for expansion. Also, the panels have compressible foam backing, and rubber washers on the front so there is further play for expansion. The bolts are not done up tightly - Nyloc bolts ensure that they are secure even if left with easy movement front to back.

Actually, I am glad they are loosely fitted, as I need to shift the far left one a centimetre to the right... this is possible. It is alarming how precise things have to be with metal and plastic. One is used to the easy tolerances of brick and timber, but metal and plastic need millimetre accuracy, like modelmaking. I had about an hour and a half re-plumbing something that I did too quickly the day before and the pipes were a centimetre too short when projecting from the wall.. it seems better to make things oversize.... (and cut back if too long).
As the technology here is a prototype, one has to be so careful and methodical. It is all modelled in ArchiCAD first - this has made it possible to make detailed cutting lists for the panels and metal sections.

Solstice sun

Well Dec 20/21 are both Sunny Days! A great time to calibrate the roof performance. Dec 20 had a short period of cloud which modified the PV harvest to 3.99 when it should have gone over 4.0.

Sunrise in Nottingham is 0816 on Dec 20/21, but the annoying thing about living in the shadow of Sharp Hill is the late sunrise in Winter. I watched the sun rise, but Peveril-sunrise didnt happen until 0918, 62 mins after the real sunrise. We get daylight in the mornings, but can only sense sunlight in the sky, but not see it on us!

GAISMA seems to be a really good site, with predictions for sun angles and declination, both in tables and in good diagrams. Plus info on insolation, wind and temperature. (the word means 'light' in Latvian)

By 0925, when the sun has fully cleared the hill, the roof is generating 310W, and 1018 an hour after Peveril-Sunrise, its getting 1240W. We pass our peak at about 1030, so at 1118 the power was 1120W..... at 1200 was 800W.... and from now it's downhill for the rest of the day :(

First Panels go up!

Dec 20: The sun shone all morning, so I got all four panels up.
Just as I finished, the sky grew dark and filled with snow, so I don't have a photo of all four, but here is one of the first panel going on.
All the panels fitted well and the biggest problem was fixing the stiff rubber hosing to join them together. The panels immediately responded to the sun, and you can see how this one curves as the front face is hot and the back face cold - when liquid flows through, the temperature will be more even.

The sunshine of the morning gave us an excellent haul of 3.99 kWhr and by the time the snow came, the sun was well round to the west. As today is the Solstice, this gives a good calibration for what the PV panels can achieve. Lets hope we have sunny days on June 20/21, Mar 20/21 and Sept 20/21.

Addition of three tracking panels?

I have in mind to increase our PV installation, but instead of more on the roof, there is room on the south wall for 3 more panels which are tracking. I could design and build the tracking frames.

I would like to know if or when there is a change in the kW rating assessment for banding of the Feed-in tariff. ie is it globally 31p if you go a step over 4KW, or if it is progressive (ie any amount over 4kW is at the lower tariff). Presently, the 36.5p per kWhr applies to systems 'nominally' under 4kW, not actually under. Its not known what happens if you install 4.4kW nominally, but because 4kW of it faces east it would never generate that amount in reality. Does the whole lot go to 31p or only the amount that exceeds 4kW?

Only three panels - 540W nominal, would generate a massive 583 kWhr in a year if tracking - eg one would change the angle every 2 months to maximise the harvest (easy to work out the perfect angle.)
The two windows open inwards, so one can simply open them and adjust the angle with bolts. Here you see a Winter, Summer and Equinox setting, from left to right.
I have written to EvoEnergy to ask. I wish I had thought of this earlier and had three fewer on the roof. I am keen enough on this idea to think that if I really am limited to 22 panels, I will have three moved from the roof to the wall.

Rails go up onto wall

19 Dec: The Forest game was at 5.20pm which was great as it left all the daylight for getting some outdoor work done. I have had the Rails finished for some days now, so was desperate to make more progress. The Eight holes were drilled, and the rails are now held firmly by anchor bolts, all in the correct location.
I have been slow because I realise the need for extreme accuracy - almost like modelmaking - the spacing of the holes on the panels gives little scope for mistakes, and I already have 8 unwanted holes that I drilled last week and then realised were in the wrong place. So accuracy is essential, even though it is time consuming. I am building a prototype, so have no precedent to work from. I have to think it out through design, and then cut or drill accordingly.
Let's hope that the Polycarbonate panels (when they arrive) are cut accurately by the suppliers and need no further trimming. If so, they will fit perfectly with minimal cutting and a lot of accurate drilling.
By the way..... Forest beat Preston 3-0 at home. I think that is the 14th unbeaten game in a row.
Another By the Way..... we seem to be weather-lucky as the forecast for the 20th-21st Dec is for a clear sky and high pressure, meaning that we hope to get a classic solar bell-curve indicating the maximum PV harvest possible at the moment of the Winter Solstice.

Making the metal rails

Dec 16: I have picked up all the aluminium angles from the University engineering stores, and have borrowed a platform drill. So the evening was spent enjoyably building the metal framework to support the panels - in freezing temperature....
I feel a sense of liberation - its been 46 yrs since I last did any metalwork, and with the right tools, it feels like a creative task - and now I am well into it, I feel less anxious about it. Also... one feels that one can make anything that one can invent - [I might make a metal gate for the house next, as it's impossible to find ready made gates that are the right size.] Many thanks to Adrian Harms for the loan of the pillar drill.

The Polycarbonate is coming soon, so that will be challenging - two glassy boxes 1.8m x 1.65m is quite an undertaking to build, and another challenge is to lift them 6 metres.

Dec 17: I have also realised that the temperature inside the polycarbonate boxes could become so excessive - perhaps 40º plus which could affect the liquid tightness of the joints, or overheat the plastics.....  - that I need to have a precaution for this by permitting Ventilation. So the bottom panel will have vents, and the top panel will become a tilting panel, with settings for Winter, Equinox and Summer, allowing appropriate amounts of ventilation. There was a chance to buy an automatic solar powered ventilator fan, but this would be automatic, so I did not get it (saving £ 80).  I prefer to try ventilation settings and see the results.
For research, I prefer to have manual settings that I can experiment with, and measure resulting temperatures, at different times of year. Using a Trombe wall based on this construction, and a MVHR (heat reclaim ventilation) you could go a long way towards heating a house using this method.

Dec 18: Still very frustrated that the Polycarbonate has not arrived... I am told it will come on Monday. So that blows the weekend I was hoping to use building the boxes.

Drilling holes for the metal structure

Dec 13: This is a view from behind, of the panels and the Lexan (polycarbonate) boxes. The holes have been drilled in the wall for all the anchor bolts. I will have 50x50 angles top and bottom, which can support the panels and the boxes. I was going to make the boxes of pure polycarbonate using 8mm with patchplate brackets, but have now ordered 6 mm (for lightness and greater transparency) and will have small Aluminium angles at the corners. I will now have the boxes completely closed (including a soffit along the bottom), and the only ventilation is around the entry/exit pipes. This will build up a good temperature inside.
I hired a diamond tipped coredrill to make the 32mm holes through the wall for the glycol pipes.
Only disadvantage of the glassy boxes is that I shall have to sense the temperature of the panels by checking the pipes in the attic, as the glassy surround will stop me reading the temperature externally with an infrared thermometer.
The same effect keeps the heat inside as the panels cannot lose heat by radiating through the polycarbonate. They will radiate back to the brick wall. I have contemplated facing the brick wall with reflective foil, but on balance there is a beneficial thermal storage effect from allowing the wall to get hot, and this will keep the space warm after sunset.
I have a difficulty picking up the metal angles, as the car is being repaired at the moment. I have the scaffolding now till early Jan, so there is time to do it well.

Dec 14: spent a while in the Faculty of Engineering stores deciding on the final cutting list requirements for the aluminium sections and the bolts - which can be collected later in the week.

Scaffolding goes up in one hour!

Dec 5: these guys don't hang about! The 30 cubic metres of scaffolding was up in a flash on Saturday morning, I had time only to make them some tea and take some photos, and it was all done!

And Emperor Scaffolding of Leicester were kind enough to leave some old planks for me to use in the Veg garden. I have the scaffolding now till the first week of Jan.
And I now feel slightly guilty that Forest hammered Leicester 5-1 on the same day! :)
(still... football is more important than Life and Death, as Bill Shankly said famously)

Electrorad offer Solar Geocharging fence

5 Dec: Here is a product that is specifically linked to solar earth charging, although its too new to appear on Electrorad's website. It's called ThermoSolar Fence. [ Electrorad.co.uk ]

As this is specifically for Heatpump, I guess my friends at Ice Energy should add it to their inventory. I would recommend it - now we are reaching the solstice, I suspect my heatpump will using increasing amounts of immersion heater instead of ground heat. This fence is another good idea, although it should not be overshaded by nearby bushes or trees.
A few metres of this, above and below ground, and not overshadowed, would probably be enough for a whole house, combining the benefits of air-solar with the battery effect of the ground - saving the cost of deep boreholes.

Thermosolar on World Architecture News
(its important to look at the smaller pictures which explain it well)
Greendiary page on the Thermosolar fence

Perfect December curve

Dec 4: We had a cold but sunny day, without a single cloud all day. So the solar harvest is short but very clean, and represented a total of 4.5kWhr, the best that a december day will do. We have a hill to the south of us, which explains why the bell curve is so steep at the start. Normally when sun rises, it is starting in a haze and the bell curve rises smoothly. In our case, the sun in December has already risen, and it peeks suddenly over the hill at full strength - so the curve is nearly vertical. After noon, the power is coming from the bright sky, not directly sunlight.
The best day in October was 13 kWh, the best in November was 7.6 and the best in December 4.5. So, of course, I am looking forward to seeing the figures improve after Dec 21, as we move to the spring and summer.
Total for October was 184 kWhr and we got 109 kWhr in November.

Red Light Immersion panic

Every so often, the heatpump seems to malfunction. I had set it so that the immersion heater should not come on, but there must be ways in which it overrides that. Today, it had a redlight moment in the afternoon (possibly not getting heat quickly enough) and then the Immersion heater comes on... it was on for FOUR hours today, more than in all the autumn. As a result the power consumption was 10 kWhr more than usual. And when it is on, the heat pump does not circulate to the ground at all.

This underlines how important it is to get on with this installation of the solar-air panels.

Scaffolding and Polycarbonate

Dec 2: I have booked the scaffolding for Saturday - for the south wall. I feel that progress is stopped until the panels are up.
Had a tutorial with Joel Carter of the MEng course on Monday, and evolved the design further - we had a thorough discussion of the benefits of putting a glassy screen infront of the panels to foster a microclimate that would increase the air temperature around the panels - a sort of Trombe wall effect.
So in the evening, i have researched the stock sizes of Polycarbonate sheet (Lexan) and worked out the optimum sizes for a sort of glass box that can surround the panels - if I leave it till another year, we would need the cost of scaffolding again, so it's worth doing all in one go.
There is an interesting discussion between me and David Brook just below this posting, discussing the idea of glassing in only one of the pairs of panels, to compare their performance.
In the image above, the South Wall panels are rendered in ArchiCAD and photoshopped onto a photo of the house and PV roof.
Postscript Dec 10th: I have now ordered the Lexan, more than 400 quids worth!! enough to make two large boxes, as above, and it will now be very closed, eg there will even be a lexan soffit going in, to make the air inside very hot. The outer leaf of the wall will get hot too, which will prolong the thermal benefit long after sunset.

Solar Rushcliffe Initiative Idea

25 Nov: On Dec 1 I am going to Rushcliffe Borough Council to present an idea about making a Solar Initiative in the borough.
I am doing a pilot study, developing a methodology for what could be a larger project, done by volunteers. I am evolving a Filemaker database structure, with the fields that would be required such as area, orientation, location. We could create a website that people could go to, to get information and register their interest.
I am using Map-Photo analysis, combining Googlemaps and Bing.com aerial photography, and Digimap vectorised OS plans to map the roofs of West Bridgford, (and a bit of local knowledge) assessing them for building type, roof angle, kilo-wattage that could be installed. Main buildings, plus streets, plus some individual houses. Using local knowledge wherever possible. Perhaps identifying 500-1000 prime targets for PV generation. Nottingham Forest, Rushcliffe CC, Notts County Hall, ASDA, the Tennis Centre, Trent Bridge CC, some of the churches, and suchlike are all good for Larger installations. But then whole streets have house that are suitable… and many streets have hipped roofs in which none of the houses are suitable.
The end result would be to have an informative website, and to write a very clear colour leaflet setting out the benefits, payback and costs, and then an individual form letter from the database to the householders declaring that their property has been identified in the research project as extremely suitable, and inviting them to ask for further details with a view to having their roof done (with a grant if before April 2010).
To keep this simple, it is focused on Solar PV – which needs clear roofs. [Solar thermal can be fitted with little 2 sqm panels on any roof, and merely supplements existing heating systems.]
PV is answering the longer term need which is to reduce coal and gas burning for grid electricity.

Big Decision about the Battery

Also, after much agonising, and much time spent on design I have had to make a decision about the battery. I have at least two students dependent on me for a dissertation now, and I am aware that the battery bring two undesirables to the table: a big Delay, and excess Complexity - both of which are bad for them, and bad for me too.

To prove my basic hypothesis that the ground can be warmed to counteract its heatloss, the best route is Route ONE, which is to take the solar and air heat directly to the boreholes.

The other reasoning behind this decision is that the Delta-T between the Solar-Air panels and the Ground is perfect. For the battery idea to work perfectly, I would really need a high temperature flat-plate solar panel, as I would have if using a large tank of water. If I did that, I might as well go the whole hog and have my original idea (of a large water tank), which would also have reduced the cost of warming the domestic hot water. With the PCM balls in the battery sticking at around 23º, the panels would not often be hot enough to send warmth to the tank - whereas when comparing with the ground temperature of 8-13ºC, the Delta-T is favourable for most of the year, even at night.

Since a part of my idea is to see if the low-cost low-temperature air panels will do the job of warming the ground, then the use of a battery or tank introduces a level of complexity that is just too confusing. I will retain the balls and consider making the battery in a year or two. (The balls can be stored in the loft until then, and it will be easy enough to plumb in the battery in two summer's time.)

More solar panels arrive!

Nov 24: Back to the Solar Geocharging.
I have been waiting a while for the remaining two panels, and they arrived today, in an Ice Energy van - so I now have all four and can get on with the next stage - scaffolding the exterior to set them on the wall and connect to the interior piping.
I will probably use the same scaffolders as did our house for the PV solar panels. Apparently a normal contract time for the scaffolding is a few weeks, so that gives enough daylight hours to get the holes drilled, etc.

Solar Surfing the Web!

The SunnyPortal website can be very revealing, if not fascinating. It leads me to think that it could lead to a dissertation topic in itself - by revealing 22 thousand sites all over the world, there are a wonderful range of systems all reporting themselves to the web - not just domestic roofs like ours, but others more revealing or experiemental. There are combination ones (so you can compare wind and solar on the same site with the graphs combined). There are some experimental structures (eg solar panels on trees of metal tubing, motorised structures, or ones that build mirrors alongside the panels.) A student with enough patience could identify many interesting installations, and then follow up by contacting the owners for more detail in the case study.

RS485 Nightmares and Web Talking

Nov 24: Our system has been talking to the web for a week now, by sending data to the SMA Website. The SunnyPortal.com compiles the data into graphs. We lost a few days of data over the weekend when the data cables pulled out of the plug when modifying the way that the unit was screwed to the wall (although the roof kept on working of course). As the colour coding of the strands was unusual (check RS485 on Google), I needed to experiment with the colours, and ask EvoEnergy a few questions. Got it working by Tuesday lunchtime after three very frustrating hours or trying.
Now I am having interesting times configuring the pages on the SunnyPortal site, because you have limited desktop publishing abilities to very the appearance of the graphs - bars, XY points, lines, polygons etc. We have also enjoyed the addition of a sensor that gives Daylight levels, surface and ambient temperatures. Sometimes, this can allow one to reflect, eg when the surface and air temperature are close the wind is less, and when a strong SW wind is blowing, the temperatures diverge.

Peveril Solar roof is on line!

Nov 17: Kevin and Dan from EvoEnergy spent a bit of time at the house completing the installation of the link from our solar PV roof to SMA's ftp site on the internet, with Dan doing the hardest work of reading the manual, and trying to work all the network settings.
The Schuco inverter and SunnyBeam transmitter box come from a German company called SMA, and now there is now a live information feed from our router to the SMA server. SMA's website then stores and re-broadcasts the info to anybody who cares to see. So here is the graph of our roof, http://tinyurl.com/peveril-solar-roof. The info is blocked to quarter hour intervals, and takes about 45 mins to be compiled to graphics (and seems to show continental time). But it's great to have a display that builds up the calender of previous days, weeks, months and more.
It has been hard to manage without the hand held unit, and it is a sorry fact that the inverter doesn't allow one to use Both the Handheld and the Web display. But we went for the Web connection, as the widely visible display and record keeping is more important.
I see that our installation gets a mention on the EvoEnergy wordpress Blog.