So, I installed and used System Advisor Model 2017 (SAM) to model my projected usage as per your questions. It helped me get away from spreadsheet estimates to a more accurate and detailed estimate:
- Performance model: Photovoltaic (detailed), or Photovoltaic (PVWatts). I used detailed.
- Financial model: Residential (dstributed)
- Location: I found my GPS location and downloaded an hourly insolation measurement from the NSRDB.
- Module: picked a CEC certified module from the list. One you're installing may be on the list, or you can create a custom module via specifications.
- mounting standoff -- relates to temperature which affects module generation efficiency
- Inverter: picked the associated inverter from the CEC database.
- System Design: I used "specify modules and inverters" as I have micro-inverters so it's one string per module
- detailed the tracking (generally fixed), tilt, azimuth configuration for each array
- Shading and Snow (optional): draw some models in 3D to calculate losses against your modules aka "active surfaces" from shading by trees, walls, roofs, etc.
- Losses: other losses you want to enter. I clicked Microinverters default losses. There's also curtailment but mess with it only if you know what you're doing.
- Lifetime: this is the expected % degradation of the panels per year. You can find the number from online spec-sheets. For mine, I put in ~0.34% for SPR-X22 360Ws
- Battery Storage: if you want to enter battery data to include for modeling. Skip this initially.
- System Costs: enter the $/Wdc things will cost. I only touched the $/Wdc per module until the total installed cost was about the installed price quote. Doing it this way lets you play with the number of modules later without having to adjust the prices.
- Financial Parameters: this is where you set the values if you want to get accurate "net present value" calculated value.
- Incentives (optional): I just enter 30% in the Federal ITC.
- Electricity Rates: this is where I entered numbers to calculate against SCE TOU rates. I couldn't get it to download accurate values from the OpenEL rate database, so I "save rate to file", and manually edited the CSV so it matched each of SCE's tou rates for affected week and weekend hours. I'm not sure if it's doing NEM 2.0 correctly here, but it seems accurate enough for one year with $0.913125 for fixed monthly charge, $10.0444 for monthly minimum, and "All generation sold at sell rate(s) and all load purchased at buy rate(s)".
- You can also enter projected rise in electrical rates per year.
- Electrical load: I used their default "pattern", but adjusted the monthly Energy "monthly energy usage edit values" to match my past annual power usage. Then clicked "Normalize supplied load profile to monthly utility bill data".
- You can also enter projected rise in electrical use per year.
After all that, click Simulate, and it'll give you a Summary that contains:
- Annual energy: how much kWh the system will generate annually
- Capacity: ratio of the system's predicted electrical output to the nameplate output (of 100% solar for 24 hours. This is the 20% @GenSao is talking about? My most optimal design (to city code) config gets me to ~20.1% capacity.
- Energy yield, Performance ratio, Levelized COE (nominal and real)
- Electricity bill without system
- Electricity bill with system, Net savings with system: these are affected by the "Electricity Rates: Monthly Accounting of Excess Generation" setting.
- Net present value
- Payback period: this seems pessimistic by 6+ months. Due to the "Monthly Accounting of Excess Generation" setting.
- Discounted payback period, Net capital cost, Equity (if you're financing), Debt
Then, there's "Data tables" where you can generate all the detailed data you ever want. You can get single values, hourly, monthly, yearly, etc values.
"Profiles": where you can graph per month all the stats you want, Heat maps, etc ...
So I used this to optimize my PV design ahead of my installer's design team. This is also helping me estimate where I should set the reserve point for PW2 settings so it it drains up to the SGIP requirement, but still maxes out my bill savings/credits vs output.
i.e. SAM is indicating for my configuration, fully draining and charging a PW2 is more wasteful and expensive than just getting to the SGIP drain requirement, and then using grid power, no matter if it's still super-peak for the last couple hours.
