First full month of new PV system generation - Cut bill by $338/mo

August was the first month that my new solar system was installed for a full utility billing cycle. The verdict is in, and I could not be happier.

I use between 80KWh and 120KWh per day depending on many things. But big factors include:

• Server Rack: Contains multiple always on computers, back up battery supplies and various electronics that are power hungry beasts usually crunching away on something. Very power hungry and can never be shut off.
• Hot Tub: Luxury item. It is a small pool that has several hundred gallons, seats 12 people, and uses a 60A 220VAC circuit. In the winter this thing pulls a lot of power. This past spring, I bought a new 8inch to 6inch tapered insulated top for it. Almost $1K worth of hot tub coverage but noticed it helped a lot. Doesn't help those night you forget to close the top... Doh!
• Nissan Leaf! My new-ish daily commuter now, gets plugged in at home, work, and when out shopping but 5100 gas-free miles later, it does contribute to the bill slightly.
• People... I have 7 people in the house. Everyone has TV's, computers, PS3's, etc. The electric laundry machine is from the 1970's and is almost running the entire day. The oven to cook everyones meals is on for the majority of the night and it just adds up

Lots of other things, but the house is wired. It is basically a smart home. Lots of electronics everywhere. Touchscreens everywhere. Electricity use is non-stop.

In June, the last month where I had only 4 of my "pirate panels" operating, I paid just under $400. ($397 and change) JUST for electric. Then about $120 in water + gas.

In August, my average NET use dropped to 15KWh a day.

The layout of my "official" panels:


Generation captured and monitored by the panels:


I had pirate panels installed for a while, but I disconnected them in January when the inspectors and utility company started coming out routinely for all their site surveys. I have re-connected them now, all 8 panels, so that is usually 12-16 KWh per day extra generation that isn't captured. These are just panels I screwed into some 2x4 supports and into Chinese grid tie inverters that plug into an outlet in the garage.

The only view of the panels is from the back of the house, down the gulley, and across the frontage road. This is a mega-zoomed in image showing my 1 neighbour as well.



This was taken BEFORE the final installation where the squirrel cages were added and all the areas tightened up and chainlinked off.



My June Bill:

(3615KWh x ($0.0711 access charge + $0.0269 supply charge + $0.0074 "Electricity Cost Adjustment, ECA" + $0.0015 Capacity Charge)) + (30days x $0.3835) = $397.95

My August Bill:

(452KWh x ($0.0711 access charge + $0.0269 supply charge + $0.0074 "Electricity Cost Adjustment, ECA" + $0.0015 Capacity Charge)) + (29days x $0.3835) = $59.44

Total Savings: $338.51

I do love my solar. These ones were a LOT more expensive than the pirate ones I put in myself. I also bought the pirate ones as "scratch and dent" and paid less than $100 per panel ($0.40/W). But to qualify for the Federal, State, and Utility rebates, strict procedures had to be followed. I applied for the installation January 1st, at 12:01am while I was sitting in the Caribbean on St. Lucia. It is a first-come-first-serve type system and you have to apply to reserve your allocation of credits. Once approved in February/March timespan and a few visits from the Utility company to the house with shade meters and inspectors I contracted a company to put them on. Paid cash which accelerated things, and they had panels in my driveway by the middle of June. Installed by the end of June. Took the city and utilities until July 9th to finally inspect and approve the whole site. That's when the main AC disconnect breaker was switched over and my solar power was pumping into the grid and reducing my footprint and bill. Long drawn out process, but totally worth it.

Installers wanted $42K for 10.1KWh installed including all labour, permits, materials, etc.
Negotiated to $37K upfront in cash.
During the install, $2K was budgeted to replace broken cedar shakes. The installers were so good, they did not even require 1 extra new shake. All excess shakes purchased were returned for a full refund (they took care of everything) and reduced my bill by $2K to a total of $35K. Since I payed in 4 milestones, the 3rd milestone was reduced.
I get back $16K from the utilities.
I get back $10.5K from Federal taxes instead of $11.1K because the cost was reduced by not having to replace the roof shakes.

Total cost is $35K - $16K - $10.5K = $8.5K

I saved $338 this month, which means that if I generate the same every month (not possible) in 26 months the system will have paid for itself 100%. Then every month it will be printing 3 Benjamin's on my roof for a good 25 years.

Now, solar doesn't output constantly throughout the year. So lets half the number for a worst worst case scenario. 4 years for total payback seems like a steal to me.

 

I bought 4 of them from some random mountain man who installs them as a living for $100/pc for 250W panels. I liked them so much I bought another 4 later. They were "scratch and dent", so they couldn't be mounted with the universal mounting systems. The metal frames were bowed or bent, shipping damage. But the cells themselves worked just fine. So I built up a wooden platform and attached them. I then tie them into individual MPPT enabled grid tie inverters. They just plug into a standard wall outlet and overdrive the mains line. Our utility company deems this as illegal. And as long as you use more than you produce the meter never spins backwards. Since without net metering it is the absolute power that they charge you for. So if you are generating 2KW into the grid on a standard meter, they CHARGE you for 2KW. If you are generating 2KW and using 2.1KW, then they only charge you for 100W which is normal.

My "pirate" setup allows me to change the angle of the panels easily by unlatching a gate lock rotating the whole 4 panels around the centre axis and locking it into another whole in the square plywood. Works great. Each array of 4 is 1KW and I have seen 1.2KW out of the system at the grid tie during a really sunny winter day.

 

Thanks!

My house is not oriented with a southern exposure. All those panels on the roof are almost directly Eastern exposure. Here in Colorado on the front range that's best anyways since those darn big piles of rock everyone calls mountains block the sun early in the west. So it is best to get great eastern production from early in the morning to mid-day, then taper off until dusk.

The 8 pirate panels are all southern pointing just because I can aim and move them where-ever. It is just 2x4's with sandbags over the supports to keep it stable in high winds.

 

Here was yesterday's production:

 

My setup uses localized DC-DC SMPS per panel, and then a single inverter. It is more efficient than the microinverters on every panel but the weakness is that there is a single point of failure at the inverter. If the inverter dies in 10 years, all panels stop working.

The panels themselves output between 0v when dark and 40V open circuit in the sun and cold. This is connected to a little micro-booster that outputs a stable 450VDC. It does MPPT on the panel it is connected to so it is always at its most efficient. When it gets below a certain threshold, it shuts off. But that generally only happens at night, even in the shade, it boosts up to the correct voltage. This is then paralleled with all the rest of the inverters through 1 high current high voltage (450VDC) connection to the inverter. It then inverts to 220VAC and feeds through a separate panel/meter into the main panel and back out the main meter.

Efficiencies for boost SMPS systems are very high and efficiencies for high load inverters can be high. So overall, I think the efficiency is slightly higher.

 

For solar it is my understanding that you can "roll it over" until 2016 tax year. So if you only have $7300 federal liability, then year 1 you get $7300 of the $10500. Then year 2 assuming you still have $7300 liability, you would get the balance of $3200. I am not a tax accountant so don't take it as gospel but I have heard that many times and a quick google seems to support it. This is different than the plugin credit.

I haven't needed to worry about it since I pay above that in taxes anyways. But I did split my Leaf purchase into the 2013 tax year and the solar into 2014 just in case.

 

I suggested it to the contractor, and he said he had installed one similar not too long ago on a MUCH smaller system, like 2KW total or so. The company I work for does some solar MPPT targeted IC's and we have lots of experience in characterizing the efficiencies of the power system from panel to wall. So for a little more cost upfront, this should produce more power overall.

 

Very true. This is also why the length of time between the start of the process and the end took so long. Lots of site inspections over a good portion of the year. They had shade analysis tools on the roof for a few weeks at various times collecting data at 4 or 5 points. They did a very thorough and complete job.

Anti-islanding is a major problem for cheap inverters. The ones I have are true anti-island inverters. If you lose main power, the grid tie shut off within 2 line cycles, 33ms as a worst case scenario. This was an important feature for me.

True.

False.

If you have a 13A load on a 15A breaker (still over the 80% recommended continuous load rating of a 15A breaker...) with no PV backfeed, the grid will supply to the panel through the house wiring to the node and deliver 13A. If you have a PV system that is generating say 14A at ~120VAC (1.7KW) onto that line, it does not jump to 27A on the wire.

If it looks like:
PV (a) ----- LOAD (b) ------ PANEL (c) ---- GRID METER (d)

then 14A will flow from (a) to (b). 13A will be sunk into the load (b). 1A will then flow from (b) to (c) and out the meter (d) causing the meter to spin backwards at 1A*120VAC

If it looks like:
LOAD (a) ------ PV (b) ------ PANEL (c) ---- GRID METER (d)

then 13A will be flowing from (b) to (a) and the remain 1A from the PV will flow from (b) to (c) to (d) and again a net of 1A*120VAC

Remember that the PV grid tie solutions do not feed at line voltage. They feed just above line voltage because they are current injectors. This means they have the highest potential. As long as there is a source in a reasonable distance it will equalize, but it will be used first. It is like having 2 linear regulators on the same circuit. If one is 3.30v and one is 3.31 volts and you are pushing 1A through it, you don't get 500mA in each of them. You get 1A through the 3.31 and 0A through the 3.30v. If you try and push it to 2A and one current limits, then you would bring down the 3.31 to 3.30 and it would be at current limit (say 1.5A) and the other 500mA will go through the now equal 3.30v.

With all that being said, your initial concern of overloading the wiring without the breaker knowing is a real and true possibility. If you take the second diagram, and change the load to 26A then you get a different problem.

26A will be sunk into the LOAD (a). 26A will flow from (b) to (a) over wiring only sized for 15A. This is BAD! Since the PV is injecting 14A, there is only 12A flowing from (c) to (b) and the meter (d) only sees 12A being used. Since the breaker in (c) is only registering 12A, everything is fine, no breaker will trip. Until a cloud passes by, the output of the PV drops to 10A and now the breaker is registering 16A draw and trips.

If done improperly it can be a hazard. If you want your insurance to pay for your own stupidity they wont. But to be fair, insurance never wants to pay out. We have had fires that burned down neighbourhoods and one of my coworkers is STILL fighting for the insurance company to pay him 3 years later. He is paying a mortgage on a property that burned down 3 years ago, rent on a house they had to move into when everything they owned burned to the ground, and the mortgage on a second place that they started building to replace the house they lost. Insurance won't pay until it is settled. Bank doesn't care it burned down, you still owe money. Either it comes via the insurance or your pocket.

Done properly, there is very little risk.

My grid ties are all fed into a single industrial power strip. This industrial power strip includes an accurate 15A single phase breaker switch. This is then plugged into a Kill-A-Watt look-alike and into my wall outlet. The only danger point for an overload on the circuit is if a larger than 15A load was connected. Generally loads are either low, high but working (EVSE charging Leaf, Hairdryer, etc) or a direct short. A direct short will trip both.

Also since I know all of these things and have control over my outlets, it is not a risk.

Battery bank for PV is not a good solution 99% of the time. The maintenance of the batteries is not worth it. The grid around here is super stable anyways. Usually once a year it goes out and is back on within minutes or at most an hour or two. We don't get hurricanes or tornadoes. Our power system was designed for heavy snow and ice and blizzards have no effect. The grid is very good around here, so that IS my battery. Having said that, my server rack runs with 2 separate UPSs and 5 additional lead acid battery banks. 99% isn't good enough for the internet.

I am thinking I will tie the extra panels into the actual PV line, or at a minimum into the AC section of that line. However that is a down the road project.

 

Another pic I found with the flange mounting. This part of the rackmount was an additional $1.8K. Because my roof is cedar shakes and not nearly flat asphalt or something, they needed to flange and seal into the shakes, then fit the shakes around the poles. The whole system is "floating".