Friday, February 14, 2014

going solar

In the Spring of 2014, having just acquired an electric car, I decided to take the next logical step and install solar panels on the roof of my house. Here are the photos from the installation.

Choosing an Installer

I talked with both Sungevity and SunPower (and SunPower's subcontractor Poco Solar).
  • From my experience, Sungevity was less expensive, with slightly less powerful panels, and was encouraged me to do a prepaid lease of the panels. A prepaid lease is a lease, with all of the payments paid up front in a lump sum. This works out to be slightly cheaper than owning, with the company responsible for maintenance. At the end of the lease period, the company can take back the panels, although in reality it seems they would be unlikely to spend money to reclaim essentially worthless panels (if the company is still around).
  • SunPower was slightly more expensive, for panels that had slightly higher output. SunPower did not discuss a lease option with me - they discussed only an ownership model.
I spent about an hour with each company discussing system size, configuration, how the net metering works with PGE, financing options, etc.  Both companies were very patient, and able to summarize what I needed to know. I had done my homework, so I was already conversant in kilowatt hours and PGE's rate plans.  After weighing the pros and cons of leases vs. ownership, and the experience of dealing with both companies (generally very positive), I went with SunPower.
  • Sungevity/Lease:  Maintenance free, less expensive, less powerful panels.
  • SunPower/Own:  No hassle of dealing with a lease if the house is sold. Higher cost, higher output panels.
Both companies had good warranties on the panels. In terms of maintenance, the only expected maintenance over the life of the panels is to replace the inverter after about 8-10 years.  Inverters run around $1,000. If I own the system, I have to replace the inverter. If I lease the system, the solar company replaces the inverter (this assumes the solar company is still in business).

Making the Purchase

Proceeding with the purchase was straightforward: I signed the one-page purchase agreement, and put down a $1,000 deposit.  The purchase was contingent on HOA approval, and PGE and city permitting.

The purchase price was $24,299, less a $1,000 discount, less a federal tax credit of $6.990, equals a net cost of $16,309. I also opted to spend $900 for monitoring equipment - essentially a box that sits next to the inverter to provide power data (with a web and app interface).


I live in a planned community (PUD) and have a homeowner's association (HOA) with an architectural review committee.  California law (very well summarized in this report) has a Solar Rights Act of 1978 that significantly limits the restrictions that HOA's can place on residents installing solar panels.  In short, an HOA can restrict an installation only if the restriction does not add 20% to the cost, or limit the power output to less than 80% of the proposed system.  I submitted a packet to the HOA consisting of a diagram of the roof provided by the solar company, with the location of the panels penciled in, and brochures describing the panels and installation process. The HOA approved.

The Installation

Soon after the purchase agreement was signed, the contractor visited and measured the roof to get a precise layout of the panels.

The system specs are twelve SunPower SPR-X20-327 modules covering 211 square feet, and one SunPower SPR 3600p-TL inverter, which results in a 3.92 kWh DC (3.47 kWh AC) size system.  The system is expected to produce an average of 6125 kilowatt hours (6.125 mWh) per year.

Because of the layout of my roof and city fire code requirements about placement of panels near the top of the roofline, there are two solar arrays - 4 panels facing west, and 8 panels facing south.

After HOA approval was secured, the installation started. The workers were amazing, explaining everything they were doing, the times they would be working, and how the components worked. There were three workers, two handling the roof work, and one handling the inverter installation and wiring.
  • Day 1: Started mounting of rails on roof, started installation of inverter and conduit.
  • Day 2: Continued mounting of rails on roof, finished installation of inverter (including connection to master circuit breaker panel), installation and painting of conduit for wiring, and installing monitoring equipment.
  • Day 3: Mounting and connecting of the solar panels. Completed installing inverter, wiring, and powering up system. Explained system to me. 

Once the installation was completed in mid-March 2014, the system was powered on, and started feeding power back into the grid. Until PGE provisioned my account for solar, I wouldn't get credit for the power I generated and fed into the grid, but it was free for me to consume at the time of generation.

The city came and did their inspection about a week after installation was completed.  The city caused some delays by requiring a different type of circuit breaker, and a harness on the roof for inspection.  The contractor negotiated with the city and the city relented, but this took an extra two weeks.  The contractor then submitted the application for Net Energy Metering (NEM) to PGE.  About a week after that, in mid-April 2014, PGE came and installed a new SmartMeter, and a few days after that, my online PGE account reflected the net energy usage (i.e., a credit for my generation) effective from the day the SmartMeter was installed.  That was it, I was up and running!

The Results

I am able to track data about the system in several ways:
  • The optional SunPower monitoring equipment feeds data into a website and a mobile app.  It's quite handy, I highly recommend it.
  • PGE provides net metering data. Basically, this means that PGE does not know how much I generate vs. how much I use - they only know on a net basis how much I am pushing to the grid at a time, or drawing off the grid at a time. If I use power that I generate, PGE has no way to know that. The amount I generate, less the amount I push to the grid, theoretically equals the amount I've used directly off the panels.
Stats (from installation on March 13, 2014 to August 5, 2014):
Total Solar KwH Generated: 2.4MW
Total Solar KwH pushed to PGE: 2.0MW
Total Solar KwH consumed: .5MW
Average Solar KwH Generated (daily / yearly): 20.6KwH / (TBD)
Max Solar KwH Generated (daily): 24.5 KwH

Wednesday, February 12, 2014

tech bubble

I've been thinking a lot about this both because my company is the public target of a lot of the protests, and because I have a lot of friends who live in San Francisco who are negatively impacted by housing costs. There is a lot of discussion within Google about the problem also, although I would say a significant part of the discussion does not seem to be practical-solution-oriented.

Here's my thoughts (only mine, not those of my company, and these very likely have their own problems and faults which I am no doubt ignorant of):

1. I'm not thrilled about subsidized housing or below market rents because it enables people to live places that they probably can't afford to live. But the reality is that these rental controls have been a big part of San Francisco's culture, economy, etc., so I respect that status quo and that's not going away.

2. Even if we end the Ellis Act (which would take a state law change), the number of absolute Ellis Act evictions is very small and is unlikely to move the needle much. There might be negative repercussions of ending Ellis Act outside of San Francisco (it's a big state and SF is not representative of landlord/tenant issues everywhere). More thought required.

3. What we can do is make it more expensive for landlords to evict rent-controlled tenants by requiring higher payouts. I think Campos (or maybe it's Wiener) has proposed this. This would discourage some evictions, and perhaps enable those who are evicted to better sustain themselves without having to leave their home city. This might change a few people's lives for the better but I don't think this is going to result in a big "fix".

4. San Francisco is going to have to get over itself and realize that a decade-long near-complete moratorium on housing development has resulted in significant problems, and those choices weren't healthy. Judging from the number of construction cranes in San Francisco it looks like City Hall is getting that message.

5. The tech companies are going to have to get over themselves and realize that they're causing a disruption for a lot of people, and this isn't just a PR problem. I think there's a lot that the employees who live in the affected areas, and the companies themselves can do, and aren't doing. This is actually one of the easier things to implement (because the cash is available). I haven't thought through a list of what it is that this should look like - it's one thing to throw money at a problem (funding community centers and such) but this is a very specific problem and taking actions that actually address the problem, rather than just PR stunts, would be important. I admit I haven't been able to think of what exactly would be the best use of resources to actually impact the problem (fixing public transit? see 5a and 6 below).

5. Private busses, including the tech companies, Academy of Art University, and UCSF, should be allowed to use the roads in the same way as any other private vehicle, but that means not stopping in MUNI stops or otherwise blocking traffic. If employees choose to live far from where they work (ugh), they can take the existing public transit to a hub, and private busses can do what they want from there. This is what I do when I walk to the Park and Ride Lot, and then take a private bus up Hwy 85 to work. Employees can take MUNI to Google's SF Office, and busses can ferry them back and forth to Mountain View from there. Risks: This might result in more cars being on the road, worse traffic for everyone, and might not move the housing problem needle. But I don't think an ever-expanding private network of busses is the answer (I've seen the bus situation on Van Ness during the evening commute and it just looks comical).

5a alternative: Change state law, and allow Muni to charge a more reasonable cost (more than $1 per stop) for using a smaller number of Muni stops (maybe a dozen or two dozen locations in the city?). Use the funds to improve Muni services (cleaning the busses more, installation of WiFi, GPS tracking, etc.).

6. Pushing more people to take public transit (see #4) will hopefully cause the city improve the public transit system (and perhaps the tech companies can help MUNI put the bus technology onto the MUNI busses). Perhaps all these companies could form a coalition to look at how BART and Caltrain are run and suggest improvements. Caltrain's annual operating budget is around $150M a year - would throwing some money, engineers, and project managers (lol) at the problem make things any better? (I don't know).

7. Wait for the bubble to burst. Because it will.

I'm sure I'm completely off-base on some of my suggestions (and I freely admit that my socio-economic situation does not enable me to appreciate the problem, and therefore work towards real solutions), but that's what's been rattling around in my head for a while.

I unfortunately think that number 7 is the likely outcome.