An ‘off grid’ journey

Sorry peeps, it’s Monday morning on 17th of October, I’ve not been out of the house yet and it’s pishing down. Darling wife dealt with the pigs and hens but with rain pelting the door and no pressing jobs. I unusually decided to do some blogging, or more correctly some notes, for this post is as much for me as anyone else and I guess the vast majority of readers will find it boring.

OK, it’s no longer Monday, it is in fact Friday and I’m in Oban, if you want the latest gos’ then try https://lifeattheendoftheroad.wordpress.com/2016/10/21/a-fair-passage-2/

Actually it’s now just after 8:00 at the bottom end of Kerrera Sound  https://en.wikipedia.org/wiki/Kerrera  just south of Oban. Spent many happy days diving around here in the late seventies early eighties. We’ll be loosing the signal shortly so I’ll post this for now.

What I intend to do is make this my reference page for the set up in our totally ‘off grid’ and virtually all electric house. For we still have a gas cooker, through choice rather than necessity. I did something similar for the last house some eight years ago https://lifeattheendoftheroad.wordpress.com/2008/01/01/powering-the-house/ but ‘things have moved on’ as they say.

After several successful years of living ‘off grid’ with smooth and reliable power of a far better quality than supplied by the local utility it became my dream to have a house that was powered totally by the elements. Not just the regular stuff like lighting and small power, but the heating and DHW too. It was a pretty tall order and needed much careful research and planning.

The ‘Energy Audit’

The first thing anyone considering living ‘off grid’ needs to do is an ‘energy audit’ that is to see just how much power you do in fact need. We already had a pretty good idea as I’d been keeping a record of what we used in the house since 2005. This didn’t include the outbuildings, workshop or chalet but it was a good starting point at only 5.0kWh per day, indeed until the arrival of the Xbox it was only 3.5kWh per day!!!!

Of course we are used to being ‘off grid’ and relying on batteries so had no electric kettle, iron and had got into the habit of only using the electric oven when it was windy, sunny, both or the generator was running. We also never have had or would have a dishwasher. Buy a bird table instead and put it outside the kitchen window. Then always do the washing up when it’s daylight, it’s quite therapeutic, keeps your hands clean and good for the environment.

  There’s an example here http://rimstar.org/renewnrg/calculate_your_power_needs.htm of what you need to make up.

Also a link to a suitable spreadsheet http://rimstar.org/renewnrg/calculate_your_power_needs/power_needs_spreadsheet.xls but basically you need to add column A (all your consumers) x B (the amount used daily) you then need to multiply this by 2 to give you the amount of useable energy you need to store to give you ‘two days autonomy’. This is the generally accepted figure for working out your minimum battery bank size. Appliances to vary greatly and this site http://www.sust-it.net/  will help you pick the best and find out more accurately what your particular appliance uses. If you live ‘off grid’ it pays to have AAAA+ if possible.

A good mix

We had already done all this prior to purchasing our 2.5kW Proven wind turbine, 1000ah 48V battery bank and 4.5kW inverter/charger. This was way back in 2005 before we had hydro and when solar panels were extremely expensive. In those days, for us at least, at this latitude with a good average wind speed a turbine was the way to go.

These days, with the price of solar PV plummeting wind is a long way behind in terms of price, reliability and predictability. However solar alone is not going to get you through a Scottish winter so a wind turbine is good to have ‘in the mix’ especially for December and January. Hydro of course is the way to go if you can but few of us are lucky enough to have a resource.

We are lucky to have all three and to be quite honest, had we not had the hydro we would most certainly have to have some kind of wood burning stove or furnace. To be quite honest, as far as calculating how much energy is required to heat and supply DHW was concerned, I was a little out of my depth. I had a good idea of how much we could produce from all the records I’d kept over the years but not really much idea of what we would need. Luckily I had much help from Alistair Robertson, formerly of Highland Ecoheat who has since passed away. This was ‘right up Alistair’s street’ and he wholeheartedly embraced the project. I provided him with the projected data and he did the calculations.

Nice graph and bar chart provided by Alistair, as you can see just a little shortfall in the winter.

Alistair’s table above breaks down the various loads into hot water, heating, house and chalet electricity. Of course the chalet burnt down in 2015 so that’s no longer in the equation which helps. You can however clearly see that heating then DHW are by far the largest consumers of energy.

These bar charts represent the same in an easier to visualize form but solar was calculated for 2kW installed rather than the 4.75kW we now have. FE wind represents the theoretical output from a Futurenergy 1kW wind turbine http://www.futurenergy.co.uk/turbine.html . I had one for a short while and thought of installing it but never did.

Of course this is far from an exact science as my data was in monthly totals which is fine for a constant source like hydro. It’s not even bad for solar which is fairly predictable but wind is another issue. Sure you can be reasonably confident on generating so many kWh per annum or even month, but it may all happen in just a few days then be calm for a week.

This is of course what the great critics of renewable energy seize hold of at every opportunity and where the battery bank and huge thermal store come in. As hot water is the main consumer it makes sense to store that rather than the electricity to actually heat it. For that I chose a 1500lt Akvaterm thermal store https://www.stovesonline.co.uk/wood_burning_stoves/Akvaterm-Thermal-Stores.html

https://lifeattheendoftheroad.wordpress.com/2015/09/06/theres-a-bomb-in-my-bunker/

Unlike a battery a thermal store can be ‘cycled’ indefinitely and contains no harmful chemicals.

So the plan was originally to have a wood boiler in the ‘Bunker’ at the back of the house that we could ‘charge’ up every few days burning large logs. However, as we were so close to actually being able to manage without we took the bold decision to rely totally on electricity. To that end we increased the PV array from 2 to 4.75kW and added 60 solar thermal evacuated tubes from Navitron http://www.navitron.org.uk/products/solar-hot-water/evacuated-tube-solar-panels

 

This probably still wouldn’t be quite enough for every eventuality but I’m always looking out for more PV as it’s ‘dirt cheap’ now. We also have another 6kW wind turbine awaiting erection but with only  75 hours generator usage in 10 months we’re pretty close already.

How it all works

The theory is quite simple, everything charges the batteries then when they’re full power is diverted to the thermal store to heat the water. If only it were that easy there are a number of decisions to be made and hurdles to overcome and here are just some of the options I looked at.

KISS

Keeping it simple is obviously the sensible route forward in many respects. By far the easiest way of doing the above is to just charge the batteries directly from the various sources then divert excess DC energy via PWM controllers to the various immersion elements.

Advantages

1 Very simple and easy to understand.

2 Better efficiency as no power conversion from DC battery voltage to mains AC.

Disadvantages

1 Only practical over short distances due to cable loss or high cost of large cable.

2 Not so easy to switch high DC currents via a regular thermostat.

3 Custom immersion elements required from the likes of http://www.tpfay.co.uk/

AC diversion

Keeping things all at mains voltage does have merits too though.

Advantages

1 Better over long distances.

2 Easier to switch.

3 You can use regular immersion elements and their thermostats.

4 Availability of cheap ‘grid tied’ equipment.

5 More efficient if ‘using when producing’.

Disadvantages

1 Complex.

2 Cheaper grid tied equipment may be cancelled out by more expensive control methods if you cannot DIY.

3 Less efficient when using stored power from the batteries.

For me, it was not an easy choice to make, indeed I actually chose the DC route initially and even ran a 70mm square 4 core armoured cable from my ‘power station’ to the bunker. It is 75m long https://www.armouredcable.net/70mm-4-core-armoured-cable.html and didn’t cost me anywhere near £18.72 per m (about £1500) but it wasn’t that cheap either.

However, in the end, after much discussion with my ‘renewable energy guru’ Hugh Piggott http://scoraigwind.co.uk/ I went for the AC option using Hugh’s clever circuit for control of two of the elements in the store.

Hugh made this circuit up for me but here are the instructions for the more capable DIYers

 http://scoraigwind.co.uk/tristar-follower-to-control-ac-heaters/

The circuit uses the Morningstar TS45 PWM controllers to drive an SSR (solid state relay) that smoothly delivers the modified AC waveform to the immersion elements.

That’s just half of the circuit.

http://www.crydom.com/en/products/catalog/m_cpc.pdf

If the immersion element thermostat opens another relay diverts to a space heater if required. If not, or if that fails then the regular DC air heater resistors ‘kick in’.

 

    

I use four Tri Stars connected to 4 x 2.7kW DC heaters (that’s two above). The Tri Stars are paired and each pair drives one SSR connected to one immersion element of 3kW in the TS.

The store has five immersion elements, the four upper ones are 3kW and the bottom 6kW. The top one is connected directly to the house consumer unit just like in any regular system, it’s controlled by a thermostat and comes on and off depending on what the stat is set at. It’s at the top of the store because that’s the hottest part and it’s where the DHW coil is. The next one down is not connected to anything just now, although it was originally connected to one of the SSR circuits. The third one is connected to an SSR circuit as is the fourth, the fourth used to be connected directly to the generator. The bottom (6kW) is connected to the generator and I’ll explain the logic to this.

Harry

My 12kW Lister is actually about double the size it needs to be, a 6kW do the job and a 7kW would do it better. However I got this old Lister for around £1000 delivered to Sconser with less than 100 hours on it. He’s a full ‘continuous duty’ set with dry sump and has been serviced annually since 1978!! In fact he had an oil heater in the sump and hadn’t been cold in 25 years. Now, the worst thing you can do to a generator is run it lightly loaded or the bores will start to glaze, combustion gasses escape and the thing will pish out and consume oil.

So, I connected a 6kW immersion directly to him so when he runs he’s always at 50% load, I then configured the Sunny Island 6.0H inverter to only draw 6kW max from the generator. This loads up the generator and heats up the store at the same time via the 6kW immersion. The 3kW immersion was also connected to the generator but was switchable. The idea behind that was to switch this on once the ‘bulk’ charging phase of battery charging was complete.

Proper charging of FLA (flooded lead acid) requires a high charge rate for a few hours (bulk) and then a period of a much lower current (absorption) for almost as long. This is the inefficient period in charging when you have the generator running but not actually doing much. This is when I planned to turn this element on using one of the relays in the SI 6.0H.

However my system is very much at the ‘suck it and see’ stage and I soon discovered that I’d far more energy spare than I thought. I also came to the conclusion that charging the batteries to 100% with the generator was daft. From memory I think the SI starts the generator at 40% SOC and stops it at 100% so I altered mine to start at 60% and stop at 80% in the hope that wind or sun would soon be along to finish off. Again, from memory I think I’ve lowered the SOC start to 50%. This means the generator only runs for a couple of hours and fully loaded. It also gives me a spare element and with the extra power available I moved one of the SSR circuits to the lower element. Obviously the lower down the store it is then the longer it takes to heat the water and thus switch off the stat.

The basic schematic

 

This was the basic set up two years ago but since then another two Tri Stars have been added

and the ‘phase control’ circuits with the SSRs

There is also the ‘Windy Boy Protection Box’

 

  

which now resides between the wind turbine and inverter. This diverts power to large resistors in the event of grid or inverter failure. The load bank resistor which is mounted in the barn is 42 Ohm and 6kW, the WBPB is that red box below the Aurora inverter. The Proven/Kingspan turbines unlike many others don’t actually mind running ‘off load’. The problem is the inverter which could get fried if the ‘open circuit voltage’ exceeded 600V. Once or twice when an MCB or RCD tripped during a gale this came very close to happening so I kept an eye out for one on eBay and got this 500V version very cheaply.

The problem with wind

As I said at the beginning, I chose wind back in 2005 because PV was extortionate, whilst wind is absolutely essential at this latitude as part of an ‘off grid’ system it is not my first choice now. I love wind turbines but as my good friend and turbine expert Hugh Piggott says :-

 “Don’t buy a small wind turbine to save money.  Do it for love, and enjoy the unique feeling but don’t expect an easy ride.”

 

Of course we had to have one as part of our mix here at Sonas and I was extremely lucky to get this one https://lifeattheendoftheroad.wordpress.com/2014/03/06/the-thousand-mile-epic/ out of a veg patch near Leeds. As is quite often the case wind turbines are sold to well meaning people by unscrupulous salesmen and put in the wrong place. This one was a classic and had never generated enough energy to even pay for its servicing let alone make any money.

The problem was that it wasn’t a battery charging model and is designed to be ‘grid tied’ but I overcame that hurdle by ‘AC coupling’ it into my own ‘mini grid’. Basically I make my own ‘mini grid’ with the SI 6.0H and install ‘grid tied’ equipment to that, any excess produced is then ‘backfed’ into the battery banks http://www.currentgeneration.co.nz/site/current/files/Partial%20AC-coupling%20in%20Minigrids.pdf

SMA and several other inverter manufacturers use this method and control the battery charging by altering the ‘Off Grid’ inverter frequency slightly. This signal then instructs the other ‘grid tied’ inverters to ‘throttle back’ production. However I’m not really a fan of this system,

1 It generally ties you into one manufacturers kit

2 It does not utilize all the potential energy that can be harvested

3 Altering the frequency can have an adverse effect on some appliances

4 It makes the addition of DC charging sources difficult

Having said that I know several people who use this system and are very happy with it. I also know another chap who has installed variable frequency relays that he uses to switch AC loads on and off.

If you do not use this or some other form of charge control though the danger is that you may damage your batteries by over charging or even worse.

You must also ensure that your total renewable AC inputs do not exceed the charging capacity of your off grid inverter. The SI6.0H that I use has a 90amp charger, so that’s around 5400W so straight away here we have a little problem. My total theoretical inputs could be double that :-

Wind 4kW all AC Coupled

Powerspout Hydro .85kW AC Coupled

Solar PV 4.75kW 2.5kW AC Coupled, 2.25kW DC

Stream engine Hydro .75kW DC

So, that’s giving me a worst (or is it best) case scenario of 7350W of ‘AC coupled’ sources that have to be dissipated before they arrive at the Sunny Island SI6.0H. Hmmmmmm, methinks I need to do more work on this In reality this is unlikely to happen, but Murphy’s law says ‘if it can it will’ and just cos it hasn’t could be down to me just wearing my lucky underpants. So far the house ‘base load’ and AC dumps have coped admirably but ‘what if’ ? Watch this space as they say!! I’m guessing it would only be possible in severe gale of north wind when the wind turbine is producing 4kW in full sun at midday with a hot thermal store. Methinks that the SI6.0H inverter would trip knocking out the grid, which would then disconnect all the GTI’s. The wind and hydro turbines would then divert to their DC resistors and the Tri Stars would dump power to their DC loads until the SI was reset, not ideal but not crucial either. Definitely needs more thought and I’m working on it.

The Aurora saga

The SMA WB6000 GTI that I’ve been using with my wind turbine works OK but I’ve never been totally happy with it, mainly because of its high start up voltage of 300VDC

 

though I have taken that down to its lowest setting of 230VDC. Even so it’s quite a breeze that’s required to get to that voltage. The Aurora 3.6 wind inverter on the other hand can start generating as low as 50VDC. OK, there’s only a tiny amount of energy at those wind speeds but if set at say 120VDC it can be worthwhile. I never set out to buy another inverter for my wind turbine, it just happened and I managed to acquire a huge and new ‘box of tricks’ that included an Aurora 3.6, Aurora wind interface, large resistive load bank, variable frequency drive, PLC and hunners of switches, fuses, disconnects and all in a large fibreglass enclosure.

Not only does the Aurora harvest at lower voltages but it also has a facility for easily entering a 16 point ‘power curve’. The SMA’s is only 5 points and the software to do it is pants. The Aurora is also transformerless so much lighter and more efficient so I thought I’d ‘give it a go’.

Using the ‘Aurora Installer’ software I started off by programming the standard KW3 power curve into it, that’s the one on the left. My turbine may be an old Proven 2.5 but it has the latest carbon fibre blades, springs and core so I figured it was a good stating point and was well pleased initially.

However, whilst the summer light winds had the turbine doing very well, it did some very strange things in higher winds and gusty conditions.

https://www.youtube.com/watch?v=xj9W_djMeb4

Its really frustrating when they do this and a proper wind up..My guess is that if it works fine in stable conditions and only cuts out in gusts, its nothing to do with the AC side. When mine was giving these problems, it cut out even when the water was turned on very slowly.

The difficult thing about the power curves is you don’t know what side of the power curve you are on. The symptoms are the same whether you are under harvesting or over harvesting. The machine just cuts out as it is doing with you. My only thought was that on your youtube clip, I see the voltage dropping quite quickly directly before it cuts out. I wonder if the ramp rate is set too high and it is coming in a bit too hard, dropping the voltage and the machine is cutting out as soon as it sees the voltage falling at that rate.

The other thing that is really important is that you use all the 16 points in the curve and give the inverter as little interpolation to do.

Many thanks to Jamie Robinson of www.alternativeengineering.co.uk for that input and Hugh Piggott for this.

hi Paul,

What I notice is that the voltage drops at the moment that the current rises. 

What this tells me is that the inverter is putting a load on the turbine 

but the turbine and cable combined impedance causes the voltage to fall.

 Again, very much ‘work in progress’  Methinks another job on the ‘to do’ list!

The second table shows the last power curve I tried but I tried many more before returning to the SMA WB6000 for the winter at least. That is the beauty of these ABB/Aurora/Power One inverters, they are really easy to program, in fact, so easy that I found it great fun to do. You just need to buy an adapter with an FTDI chip, it’s RS485 to USB, http://www.ebay.co.uk/itm/USB-to-RS485-FTDI-interface-Board-Power-One-Aurora-Inverter-Web-Data-Logger-/281238492352?hash=item417b1ed4c0:g:qjgAAMXQiw5RYo58  open the Aurora Installer software and enter the installer password which is 05591This will get you into the settings for the power curve. Unlike the SMA wind inverters that come loaded with something that will at least get you ‘up and running’ the Aurora won’t actually function until you enter one! You cannot alter any of the grid settings with this code so there’s no need to worry about accidentally doing that. To alter those you need a unique code from ABB and I guess you have to make a strong case for getting it off them. I managed to get the code for mine, which is actually an Irish one and configured to their EN50438 spec by convincing them it was to be used off grid only.

EN50438 brief summary

As you can see the specs are broadly similar but there’s one crucial difference that’s not shown on the table. The G83/2 spec requires a 3 minute wait after a failure before it can reconnect. As far as I’m aware (or at least mine does) the Irish ones can reconnect after 20 seconds. Now this may not seem like a huge difference but with the SMA WB6000 and it’s high start up voltage it’s really infuriating cos every time it disconnects due to low wind, it takes ages to reconnect as it has to go through all its system checks!!

 

22/10/16

Another snag with ‘AC Coupling’ and the use of GTI’s is that you need a really stable generator that produces ‘grid quality’ power. GTI’s have to disconnect from the grid if it fails, the parameters for this vary with each country but the UK is covered by something called G83/2 and this dictates how quickly and under what conditions it should disconnect and the time elapsed before it reconnects. This is to prevent power being fed into what should be a ‘dead’ line and electrocuting someone. Some inverters are designed to be used in ‘Island type grids’ and have a setting for that purpose requiring the use of a unique password. This will still turn the inverter off it it loses the island grid but won’t do it quite so quickly and the parameters for under/over voltage or frequency will be wider.

I have run several inverters with and without the ‘island’ setting turned on on my mini grid. I found that the SMA inverters worked OK with it in standard G83/2 mode until a large inductive load was used but they were fine with the generators. Once the Island mode was turned on they rarely, if ever ‘drop the grid’. The Solis 1500W inverter http://www.ginlong.com/en/PV_Inverters/Solis_mini_1500.html  that I’m testing right now for Powerspout http://www.powerspout.com/  and Hugh Piggott http://scoraigwind.co.uk/  works very well with both inductive loads and the generator even in its G83/2 setting. The Aurora however was not so happy, even after altering the parameters via the unique password.

23/10/16

I guess this is becoming a little random and chaotic, but as I said, it was more for me to record stuff and ‘think out loud’.

So, on a sunny day steaming up the east side of Arran I’ll return to the Aurora inverter. It would seem that I’m not alone in having issues with it in an ‘off grid’ and ‘AC coupled’ system. Jamie Robinson of www.alternativeengineering.co.uk on the Knoydart peninsula was also struggling with his. He was using an ABB/Aurora PVI 3.6-OUTD W same as mine with a Powerspout hydro turbine and had a really bizarre issue that had both himself and Hugh Piggott ‘tearing their hair out’. Like mine it kept cutting out and loosing the grid but his was only doing it in good weather and rarely at night. After much much head scratching and innumerable alterations of his power curve he eventually discovered that it was only doing it when the sun came out!!! It turned out to be a conflict between the Aurora and the micro inverters on his ‘AC Coupled’ solar panels!!!

I’m no electrician but I suspect this type of issue to be more common on transformerless inverters than ones that have a transformer as there is no isolation between incoming DC and outgoing AC. I ruled that out on mine by disconnecting all the other inverters and the phase control circuits, still didn’t help. Whilst it would appear that I’ve had a lot of issues with this inverter in my system I still rate it highly, I just think it needs a bit more ‘tweaking’. As it is I lost patience with it and reverted back to the SMA just to get me back harvesting energy again. I’m sure I’ll revert back to it at some point, even if it’s just to try it out with the Powerspout hydro turbine and compare notes with Jamie.

I guess this is another disadvantage of ‘AC coupling’ in an ‘off grid’ system, it’s not always a smooth ride. Having said that I’ve had plenty of help along the way, particularly from Rob Beckers of Solacity http://www.solacity.com/ in Canada, don’t you just love the Internet

Update 11/01/17

That’s the New Year ‘in with a bang’, storm Barbara and storm Connor passed without incident and the SMA WB6000 has performed very well without any glitches. I started work on the foundation for the 6kW turbine and am now awaiting the 12 cube of concrete for the base.

 

 

https://lifeattheendoftheroad.wordpress.com/2017/01/05/i-love-the-smell-of-napalm-in-the-morning/

The problem is that I’ve not actually worked out yet what to do with the energy! As this too is a ‘grid tied’ turbine my options are somewhat limited. Golden rule when ‘AC coupling’ is not to exceed the capacity of the inverters battery charger and I’m over that already So adding an extra 6kW is a recipe for the ‘magic smoke’.  I am thinking of somehow utilizing my spare 950Ah 48V bank and the Outback GVFX 3048e but not sure how. My good mate Scruff of Navitron Renewable Energy and Sustainability Forum – Index has got me thinking along these lines.

Enlarged from PDF

 

 

Much as I dislike the idea of having two battery banks and their associated problems, double the maintenance, more cable losses, incompatible voltages etc. This does seem like a solution to the dilemma, in part at least. The 950Ah bank works at a slightly lower voltage and the big diode only allows current to flow from the 950Ah forklift cells to the main Rolls 800Ah bank in the ‘power station’. So I’m thinking of having a Morningstar T60 in ‘diversion’ mode at say 55V to an immersion in the TS so when the main bank voltage is below that plus the .7V lost in the cables and diode current will flow from the secondary bank to the primary, maybe As I said ‘work in progress’ and I still need to work around having a 6kW turbine and a 3kW inverter/charger.

Further Progress April 2016

A month long ‘holiday’ saw me finally getting the 15m Hutchinson tower erected with the 6kW turbine on top https://lifeattheendoftheroad.wordpress.com/2017/04/14/shes-flying-at-last/ It was not without its problems though. By right’s this size of mast needs lifting with a 3.2t Tirfor using two anchor points and a pulley mounted atop the ‘gin pole’. The 3.2t Tirfor I have but the 16mm wire is only 20m long.

However, most of the weight is in the first few degrees of lift so I used my Ifor Williams 2t tipping trailer to assist.

 

This worked really well until the trailer hydraulics jammed in the up position After that it was kinda hard work but we got it up in the end.

At the moment though it’s still not actually connected to the house grid and is sitting there with the brake on, ‘work in progress’ as they say