Experiments in off-grid solar – Phase 2

It took me awhile, but I finally got back to my off-grid solar project. Phase 1, which was the initial build, proved very successful with the ability to keep the three batteries we have (two 75 ah batteries and one 35 ah batteries) charged and usable. We used these batteries in two portable charger carts each with a 1000 watt modified sine wave inverter to run the TV/DVD player as well as miscellaneous other uses through out the house/yard. The biggest downside to Phase 1 was that because we always had to bring the batteries/inverter to the power source, I was forced to lug the batteries around which got old pretty fast. This is the problem phase 2 would solve.

The goal of Phase 2
Phase 2 of the experiment was to add a sine wave inverter that would allow utility-grade electricity to flow through the house wiring through a small transfer switch. By doing this we would centralize our off-grid system components in the basement while still getting power where we wanted it.

Steps involved
The first step which happened earlier this year was to install a transfer switch allowing us to control 4 circuits in the house from the off-grid solar. Due to the small size of the batteries and scope of this project, we needed to run small loads so things like the well-pump were immediately excluded. It was decided to try to run the office (lights and plugs), the basement (lights and plugs), as well as a plug or two in the kitchen. We knew that we would not have enough power to run all of these at once, or even one of these circuits 24/7, but it gave us an obtainable goal to shoot for in future phases of this project. By selecting these four circuits we also gained the ability to have at least one outlet on every floor of the house that could be powered off-grid. This step was successfully completed and unit tested successfully in the spring of 2009.

The next step was the inverter selection. There are quite a few choices of off-grid inverters out there, but not a lot on the low end, and at this point I didn’t want to break the bank on a largely oversized inverter that we rarely used. I initially selected the Xantrex ProWatt 2000 SW which was a 2000 watt pure sine wave inverter that was fairly new to the market but was getting really good reviews. It had two GFCI outlets as output so I figured I could plug the transfer switch into one of the outlets and be good to go. Well, here I learned another lesson of low-end inverters. The ones that only have GFCI outlets (i.e. the ones that are really portable), cannot be tied into your house wiring because they are grounded and wired wrong. They don’t ground neutral to ground which is required in house wiring and therefore every time I tried to flip the transfer switch, I tripped the GFCI outlet breaker. Well, that sent me back online searching for a different inverter that I could use. I ended up the older technology Xantrex Prosign 1800 inverter that allowed for hardwiring the AC output and did ground neutral to ground thus being compatible with the house wiring. A quick unit test to prove that the inverter would work the way it should was performed in the fall of 2009.

Now that we knew all the major components, layout of these components became an issue. It quickly became clear that the batteries would have to be moved from on top of the table to on the floor in a box which meant building a battery box. In phase one, the two 75 ah batteries were being use separately even though they were the same age. This worked well with the portable carts, but now that the batteries were going to stay in the basement it made more sense to tie them together to increase the one time capacity to 150 ah. To allow for future expansion, I build the box to allow for four batteries, plus a smaller fifth battery. Since I also wanted the ability to isolate the batteries from the rest of the system, I included in the design a battery switch which not only allows for cutting the batteries off from the rest of the system, but also allowed for multiple battery banks. I didn’t see a need for multiple battery banks at this stage but I figured I could plug the smaller pump battery into the second slot on the switch to keep it charged but separated from the primary bank. The box was made of scrap wood I had in the shop and was completed in the fall of 2009.

After completing the battery box, it was time to move on to the wiring and rewiring of the system. In phase one the system consisted of a PV combiner box, a small breaker box and the charge controller. The PV combiner box and charge controller would remain, but the small breaker box had to be replaced with a larger breaker box to house all the expanded wiring and connections as well as the large breaker needed for the inverter. For this task, I opted for a Midnight Solar DC breaker box since it had just enough room to accommodate everything I needed including the 250 amp inverter breaker, 5 slots for other breakers (of which I only needed 4), and a DC shunt for all negative DC connections. Everything fit nicely in the breaker box although the 4/0 cables I needed for the inverter and from the battery box did prove a little challenging bend through the conduit. I was able to reuse the PV and battery breakers I had from phase 1 and added a ground fault protection breaker as well. The wiring for the batteries and charge controller was completed first and brought online late in the fall of 2009.

The last step was the installation of the inverter itself. With the breaker box set, the inverter went up fairly smoothly although because of the lack of flexibility in the 4/0 cables, the placement of the inverter had to be tweaked a couple times (thanks to my wonderful wife for understanding why we needed to “move it again!”). Once the DC wiring was completed, the AC wiring went really quickly especially since the wires were so much easier to deal with then the monster 4/0 cables on the DC side. This wiring was completed in November of 2009.

Completion
After double checking all the connections, we powered up the inverter for the first time and allowed it to run the internet modem / router as well as one laptop computer for a couple hours. There were no issues found and it all seemed to be working. We have since, plugged in the fridge to the inverter just to see how long it would run on the batteries, and we got 11.5 hours of run time off of it which isn’t too bad. The fridge is not a device we would normally run off-grid on this system but it was a good heavy load to really give the system a solid test.

Next Steps
So, where do we go from here? We first need to see how much we can use the off-grid system as is. The ideal hope would be that we could run the office 2 days a week. That may be a little unrealistic in the winter with the low amount of sun, but we will take what we can get and see how well we do.

Thoughts on Phase 3
Of course in this on-going experiment as soon as the build for one phase ends we need to be thinking about the next phase. Currently we are thinking phase 3 will be to increase our charge capacity and to move the PV array to a more optimum location. During the ice storm last year, one of the two PV panels we have was damaged and is producing at 1/10th rated capacity, so we are really running off of one PV panel at this point. The goal of phase 3 at this point would be to replace the broken PV panel and add two additional panels bringing the total to four operational panels. Since most of this work is outside work and it is currently only 30+ degrees outside, this phase will have to wait until spring at earliest.