NJ solar - can you doublecheck this excel model?

[petesamprs]

Here is my excel model (downloadable): https://drive.google.com/file/d/17TW...YMtvc9i7Y/view

I'm about to pull the trigger on solar for a house that I recently purchased in northern NJ. Back of house faces 180^ south with zero shade all day.

Unfortunately I don't have any usage data at the new house (5k sq ft, 3 central ACs, 18x40 pool), only at my current house in southern NY which is much smaller (2700 sq ft, 1 central AC, no pool). So my need is an estimate that takes into account my current usage of ~1200/avg month, adjusted higher.

I prepared the above cash flow model which is meant to capture all of the economic drivers of the decision and calculate NPV/IRR over time. Can you check my assumptions and see if they are reasonable?

I have a few remaining questions after conversations with my very helpful local (not big name) solar salesperson:
- How should I decide between a Panasonic panel @ $2.95/watt (25 yr warranty) vs. a Trina Solar panel @ $2.54/watt (10 yr warranty)? Is this really a commodity product where price should drive the decision? Note: Roof area is not a constraint.
- My salesperson is calculating a massively different number than PVWatts for what my panels will generate. PVWatts says 1KW will generate 1,421 KWH/year (assumes premium panel, 10% loss, 30^ tilt, 180^ azimuth), whereas salesperson says 1,190. This results in them sizing a system 20%(!!) bigger than what PVWatts would suggest, which adds 1-2 years to breakeven calc. What would explain that huge difference? Do they have access to better models than PVWatts?
- Word is that the threshold to generate SREC will soon increase to 1200 KWH from 1000 KWH, so that's what I'm modeling in. Anyone else hear that?

Overall the math seems really compelling so I want to make sure i'm not missing anything in my analysis.

thanks

[J.P.M.]

I'm getting ~ 1,400 kWh/yr per STC kW for Hackettstown. Reliable models tend to show the same results with the same input. PVWatts usually agrees with other models and has pretty much proved itself a good tool for preliminary design. Unless there's something different in model inputs, or something about your application that PVWatts doesn't consider but other models do consider,, other models ought to generate ~ = output +/- 5-10 % or less. Peddlers around my area generally come in with model output ~ 10 % or so beyond what to expect. No peddler ever got fired for oversizing by underestimating productivity. With no shade and due south, I'd go w/ the PVWatts #'s and add a couple of panels to handle performance degradation of maybe 0.5%/yr.

Panels are pretty much a commodity. At this time it looks like they fail due to infant mortality and not too much else. The installation quality and knowledge/professionalism/integrity of the installer are more important even if harder to verify.

I'm ignorant of the SREC situation in NJ.