Here are the costs of the system in original form; I kept most of the receipts rather than rely on memory, so the numbers should be accurate:

Looking at the final overall figure, that is an increase of 25% on the original system cost in its raw delivered state. So when doing a DIY renewable energy project, a rule of thumb may be to add 20% to the cost before you plan your project, to ensure your estimates are realistic. Of course, I could have saved more by reducing the solar panel cable runs, or by only putting in 3 turbine wall brackets instead of 4, and so on. And if you already had a suitable housing for your renewable store room, then even better, as that did constitute a large part of the extra cost.

Here are the costs of the system additions in order to bring the system up to 1.6Kw. 

Battery Usage Chart Q1/Q2 2007:

Going back to the figures from the battery usage estimates, I was expecting to get around 27hrs of usage if the battery bank started at 80% capacity and was discharged to 50% of the total amps available. I was running my PC and monitor from my study, which draws about 150watts. So far the real usage calculations are indicating that when used for 10 hours continuous, from a starting point of 100%, that the batteries have depleted to almost 40%. So in the Q1 table below (Jan-Mar) I appear to be achieving less than one third the expected return from the batteries!  

Now there could be a number of factors for this, such as the condition of the batteries in the colder winter months, or perhaps the inverter is not running efficiently. Also, perhaps my readings from the battery while they are use may not be returning a true indication of the power remaining. Or could it be a faulty battery, or were my original battery usage calculations incorrect? Well, I tested again during Q2 2007, and now have better results, mainly because a) the temperature has increased by 5-6 degrees, which enables the battery to provide more power. Also, unlike the first test, I did not take the final reading until about 1.5hrs after I had finished drawing power:

Upgraded Solar Controller, With Remote Monitoring:

Wouldn't it be great if I was able to monitor the status of the system from the house, rather than have to head outside and run frequent battery checks using my meter, to ensure that I was not dropping below the 50% mark? Well, I mentioned on the previous page that I upgraded the controller to a more sophisticated model. The version I chose has all the usual solar functions, such as pulse width modulation and battery overcharge protection. It also capable of equalizing the batteries on a regular basis to ensure they stay in the best condition. The controller is manufactured by Morningstar, and the model is a Tri-Star 45, as pictured below. It also has standalone battery voltage sense wires and temperature sense wires. The level of battery charge is altered depending on the temperature of the battery environment. 

For this controller you can buy the LCD display to fits into the centre of the controller, or you can buy a remote version that works using an RJ12 cable, or you can have both! For my purposes I only needed the battery state to be available in the house, so the picture on the right shows the remote meter in use. The remote meter comes with 25metres of RJ12 cable, but this can be extended by adding additional RJ12 cable, without any impact. For those not familiar with RJ12, this is the same as the better known RJ11 cable but it has all 6 points connected, rather than just 2 or 4. I used the conduit between the garage and house to run the main length of cable, and added 2 smaller 10m and 15m cables at either end, using RJ12 couplers,  to give me the distance required.    

The manuals that come with the controller and remote display give plenty of information on setting and changing the system, including an equalisation cycle which manages the batteries automatically. But the best part is being able to see, at a glance, the exact battery voltage and solar amps being fed into the batteries. This way I can then choose when to swap back from renewable power to GRID power, should the batteries run too low. The remote display is also backlit to aid reading in poor light conditions. One final point: there is a larger brother of the Tri-Star 45 called the Tri-Star 60, which handles up to 60 amps of solar power. Each of the solar panels currently outputs 100watts / 12V = 8amps. So when working with 100 watt panels a maximum of 5 panels can be used with the Tri-Star 45, or 7 panels if using the 60 amp version.

Inverter Efficiency Test:

Assuming that my batteries are in a healthy state (I would have no reason to suspect otherwise), I wanted to test the  efficiency of the modified sine wave (msw) inverter, compared to the pure sine wave (psw) inverter. It is very difficult to obtain efficiency charts from the various inverter manufacturers, so I thought a better way to do this would be to run a controlled experiment to see how the two inverters compare. For this test I used three small table lamps with incandescent 60watt bulbs, all linked to a single extension cable plugged into the inverter.
 

If doing any testing yourself you need to ensure that the starting voltage of the batteries is stable. Following charging it can take up to 3-4 hours before you get a true stable voltage. I disconnected the wind and solar inputs overnight and waited until I had an equal starting voltage in each test. Also, remember that the final voltage shown after 3hrs of load being drawn is not representative of the battery power remaining; you need to wait at least 1.5 hours more before taking a reading, as evidenced by the differences in the final two columns.

I was surprised when the 3KW psw inverter came up with the same figures as the much smaller msw inverter. I would have expected the msw to do much better here, given that the load requirement is within that inverter's peak efficiency range, i.e. 180watts falls between 20% and 80% of 600watts. In the past, modified sine inverters were considered to deliver better efficiencies than the pure sine versions. But given these results it appears that the gap has narrowed. Finally, I did test No 3, this time plugging the lamps into the house sockets to see if running the power all the way through the SWA cable, via the new Consumer Unit, had any adverse impact on the readings. There was no difference, even allowing for the fact that 180watts is outside the 3Kw inverter peak efficiency range. 

Coming next : investigation into grid-tie systems, which are beginning to become available in Ireland. Perhaps that will form the basis for my next project!