Panel Installation, Part 2

The Next Morning

Early the next morning we were off to Nuts & Bolts Supply who could supply us with the proper stainless hardware, then back home to replace those 42 mid-clips and bolts which, as you have guessed by now, required sliding everything off, replacing the bolts, and sliding everything back on. The new wrinkle this time was that over the night moisture had condensed on the rails, in the track, and had frozen. It took some effort to get the old bolts off.

Things started to move pretty smoothly. A neighbor had very kindly loaned us a scissor lift which made the job oh so much easier.  Once we got the first nine panels in place we tested the 'A' string with a voltmeter. The meter read a very pleasing 387 VDC!

Note that it is very important to use a proper torque wrench when tightening the end-clip and mid-clip bolts. Exceeding the recommended torque will easily damage the panels by cracking the tempered glass. It's also critical to avoid damage the back of the panels since it consists of a thin coating over the back of the individual cells and scratching of that coating would be a serious matter.

As each string of 8 panels was completed we tested the output voltage to make sure we were getting the desired output. We finished mounting the panels just as the sun set (and things started to get really cold) and it was not possible to test the final string since there was no longer adequate light.

Putting that last panel up was very satisfying. I tried to give my son a hug, he went on his way, and I cleaned up all of the cardboard from the panels while collecting the serial number stickers for my records.

The next morning, after the sun came up, I went out and measured the voltage of the final 'C' string and double-checked the 'B' and 'A' strings.  'C' and 'A' was fine but 'B' was measuring only 334 VDC, indicating that one panel was not producing. Since I was getting some voltage I concluded that the Tigo Maximizers were properly wired to one another. The problem had to be either a bad connection between one panel's output into its Tigo Maximizer or a Maximizer was flaky.

The best way to identify the problem was by using the Tigo MMU (Maximizer Management Unit). Tigo had already set up my account at their web site so that once I hooked up the MMU to my Internet service I could go to that web site and see how each panel was performing.

There was some confusion on my part, mostly due to inadequate documentation of the MMU and how the Tigo products worked — but this was a brand new product and I had expected some 'bleeding' edge experiences along the way.

The installation manual for the MMU did not explain how to hook up the sensor/gateway cable. The sensor/gateway is a small box that wirelessly communicates with the 24 Tigo Maximizers (LMUs) which we had earlier mounted to the back of each solar module. I dashed off an email to James Bickford at Tigo and very quickly had a Tigo engineer on the phone helping me get things set up.  The engineer also explained that it was not necessary for the solar modules to be connected to an inverter in order to determine how each module was performing.

After some initial miscommunication, I got the sensor/gateway all cabled up, the MMU connected to the home's ethernet, and powered up. By this time the sun was going down and it was futile to see if there was any output coming form the panels.

Panel Installation

We hired a couple of roofers to mount the rails. They did a fine job and the cost was reasonable.

The key when installing the rails is to insure that the L-feet are firmly lagged into the trusses.  Roofers are good at that, though one does end up having a few extra holes in the roof — nicely sealed, of course, or, in roofer-speak, 'gooped'. Chalk-lines helped keep things aligned.

The rails are mounted horizontally so that they will support the solar modules, which are 'portrait' oriented: the longest dimension of the module is vertical. The L-feet are no more than 48" apart.

The weather was starting to get cold. Because the wires from each Tigo Maximizer run between the panels (as shown in the previous post) we put adhesive cable tie holders at strategic locations on each module. It was so cold that evening that the adhesive wouldn't stick so I had to warm up the aluminum frame a bit and slap on the cable tie holder. It took a couple of hours.

A few days later my son and I started mounting panels.  Earlier in the project I was a bit reluctant to scramble around on the roof but the rails made getting around a lot easier. First thing was to slide the mounting bolts and clips into the track of each rail: 42 mid-clips and bolts, 12 end-clips and bolts, and six grounding straps over the rail splices.

We forgot the grounding wire clips! Slide off about half the clips and bolts, slide on the grounding wire clips and slide back on the other clips and bolts.  Ready.

We mounted the first panel. Almost. The end clips went on just fine and held the panel pretty firmly but we were anxious to get the other side of the panel secured. It would not be good if the panel came loose and slid down the roof. That's when we discovered that the 2-1/4" stainless bolts used to secure the mid-clips were too short! Big mistake by Wholesale Solar! We needed 2-1/2" bolts! So it was a mad dash off to the local hardware store and pick up a bag of 50 bolts.

Slide off 42 mid-clips and bolts, six end clips and bolts, the six grounding wire clips, and the six grounding straps.  Replace the mid-clip bolts and slide on 42 mid-clips and bolts, six end clips and bolts, and the six grounding wire clips.  Forgot the six grounding straps.  Slide, rinse, repeat.

We mounted the two panels immediately below the original panel, connecting the Tigo maximizer cables, and called it a day.

During the evening, as I was checking and double-checking everything, I realized that the new 2-1/2" bolts we had installed were galvanized, not stainless steel. This was not good — the galvanized hardware would have a galvanic reaction to the aluminum rails causing the metal to degrade. Sigh.

Tigo Maximizers & Module Preparation

Tigo Maximizers

Through some heroic efforts on the part of Tigo Energy, DC Power Systems, and Affordable Solar, I was able to acquire 24 ea. Tigo Module Maximizers (MM-ES060V300W-4RL), 1 ea. Module Management Unit (MU-ESW), 1 ea. wireless transceiver (which communicates with the Maximizers), and some odds and ends.

As mentioned in an earlier post, there is one Maximizer for each solar module.  In fact, the Maximizer attaches to the back of the module.  Its purpose is to balance the DC output of the panel in order to reduce the effect of shading and other module inconsistencies on the whole string of modules.  Without the Maximizer, one shaded panel would seriously degrade the production of the entire string.

The Tigo Maximizer product is brand new and is only now becoming available to the U.S. solar PV market.  Their products will initially be available through two distributors: AEE Solar and DC Power Systems.

Because AEE Solar has a local distributor, I originally attempted to work through them in acquiring this brand new Tigo product.  Unfortunately, they were uninterested.  DC Power Systems, though, were very anxious to help get me the Maximizers in time to meet my deadline (of being on-line by EOY).

I cannot say enough positive about the fine folks at Tigo Energy and, in particular, their new Marketing Manager, James Bickford!  James has tirelessly pushed internally and externally to help me acquire their Maximizers in time.  James has called and emailed numerous times to insure that things were 'in motion' and, once I'd received the product, to make sure any questions or issues I might have were being addressed.

Kudos to Tigo Energy, James Bickford and many others at Tigo Energy, DC Power Systems, and Affordable Solar, a retailer for DC Power Systems.

Module Preparation

Each Tigo Maximizer comes pre-attached to a mounting plate with a clamping assembly along two edges.  The intention is to loosen the clamps, place the maximizer in the corner of the solar module, and tighten the clamps onto the framing of the module.

Unfortunately, the Sanyo modules have a frame edge to which the Maximizer can be clamped only along one edge.  Single-edge clamping is inadequate inasmuch as any vibration at all causes the clamp to loosen.  Our solution was to drill two holes in the frame edge and then directly mount the Maximizer to the Sanyo frame:

 

Of course, it is important to place the Maximizer in one of the corners adjacent to the cable connections of the solar module.

Note that we waited to connect the cables from the module to the Maximizer until just before module installation.  Also, some Maximizers were attached to the left of the panel connection and some to the right.  This was done so that the Maximizer cables could best connect to the next and previous Maximizer in the series according to the following diagram:

Mounting Rail Reprise

For mounting the solar modules onto the roof I used Ironridge Solar Mount rails.  These are heavy-duty aluminum rails with channels into which bolts are inserted for the L-shaped mounts and the module clips, one channel for each:

Ironridge Solar Roof Mount System Rail Cross-section

The channel on the top is used for inserting bolts for:

• solar module mounting clips
• grounding straps straddling joints in the rails
• grounding wire clips for running the grounding wire from rail to rail and to groung

The channel on the side at the bottom is used for attaching to the L-mount which is then lagged into the roof.

Since this is aluminum it is very important to use stainless hardware.  If galvanized (zinc-plated steel) bolts, washers, etc. are used then a galvanic reaction will occur and the aluminum will be weakened.

So far so good.

Attaching the rails to the roof was straightforward:

• locate the trusses
• lay down chalk lines
• pre-drill holes for the lag screws
• goop up the lag screws and the bottom of an L-mount
• screw the L-mount into the truss
• slide the mounting bolts into the lower channel of a rail
• line up mounting bolts with the L-mount and insert
• apply the washer, if required, and attach the nut
• tighten with a torque wrench according to specifications

This part of the mounting was pretty much trouble-free.  Placing the panels was an 'experience'.

Solar Modules & Inverter Arrive

Parts start to arrive!  First in was a palette of panels:

24 Sanyo HIT Power N 210N/HIP-210NKHA5 Solar Modules

Next up, the inverter:

Fronius IG Plus 5.0

And some miscellaneous hardware:

Miscellaneous Hardware

The miscellaneous hardware includes the AC disconnect, fuses, grounding straps, mounting clips, etc.  [The rails had actually arrived after the panels, inverter and miscellaneous hardware, but they were discussed in the previous post.]

All of these components, including the rails, were ordered from Wholesale Solar.  The folks at Wholesale Solar were very helpful and their prices are mostly excellent, but you must know what you want — n00bs are likely to end up with the wrong equipment.

Be careful!  Two examples:

• When asking for my first quote I informed Wholesale Solar that I would have 3 strings of 8 Sanyo modules and "appropriate rails" and explained the arrangement of the modules.  The rails they proposed would work well for an arrangement of 3 rows of 8 modules, but it would not work well for the arrangement I described for them: 1 row of 9 modules, 1 row of 8, and 1 of 7.  
• Their price on the solar modules, inverter and rails were very competitive, but their prices for other components such as fuses, disconnects, and cables were actually higher than MSRP.  Even though I pointed out such discrepancies to them and they promised to check into it there was no adjustment made and I eliminated some of those components in favor of purchasing them locally.

So always check things over carefully, regardless of from whom you may purchase the various pieces parts.

Rail Installation

The rails upon which the solar modules will be mounted were installed this week.  Here is the before:

And here is the after:

The rails are heavy duty aluminum.  L-shaped mounts attach to the roof every 4 feet.  The L-feet are attached to the roof by lag screw driven into a truss.  Sealant is applied to the lag screw and to the bottom of each L.  Altogether, it took about 5 hours to install the rails with two roofers, one rail assemblyman (me), and one 'supervisor'.  (That's a 40° roof!  There was no way I was going to get up there.

The conduit for the DC power leads will run just under the edge of the roof in the soffit.

Panel Optimization Part 3

Last week, while in the San Francisco Bay area, I was able to stop by the Tigo Energy offices in Los Gatos, CA and visit with their very pleasant marketing manager and ask many questions about the Tigo Energy Module Maximizer (EMM)

The EMM is a very simple device which clip on the back of each solar module.  The power leads of the solar module are connected to the EMM.  Separate, and different, leads daisy chain from one EMM to another.  Each EMM has a built-in thermocouple.  The EMM essentially insures that the voltage output from the solar module is kept high enough to prevent series degradation.  Without the EMM, a solar module in shade would develop a high resistance and seriously degrade the entire series of solar modules.

Each maximizer communicates with a central Maximizer Management Unit (MMU).  The MMU can be used to control each EMM (to enable or disable one, for instance).  The MMU also collects:

• the performance and temperature of the solar module attached to each EMM,
• solar radiance information from a separate (optional) pyranometer, and
• AC power generation from the inverter (also optional).

The communications between the EMMs and the MMU can be either wired or wireless.  The wired version uses the DC power lines.  This option requires that the MMU be physically connected to the DC power lines at a point after the individual series of solar modules have been 'combined' into a single DC feed.  This is usually just before the individual series DC feeds enter the inverter.

The wireless version of the MMU and EMMs do not require the MMU to be physically connected to the DC feed, so the MMU can be placed anywhere within a reasonable distance of the solar modules, up to 300 feet (90 meters) away.

In order for the MMU to record the AC power generation of the inverter, the MMU will need to be mounted somewhere between the inverter and the power panel.

Availability is an issue.  Supposedly, 60 EMMs and a number of matching MMUs were to be shipped out to AEE Solar in Redway, CA this week.  I need the following:

• 24 EMMs, wired, 72" power leads, MC4 connections for the solar panels
• 1 MMU, wired, with AC metering
• 1 pyranometer (Licor)

Armed with the exact information I needed to place an order, I contacted AEE Solar's CEO, David Katz.  I'd spoken with Mr. Katz several weeks ago and he had assured me that AEE would help meet my need.  Now that I knew the parts that I needed I passed that information on to Mr. Katz as well as to our local AEE Solar distributor, Orrin Farnsworth.  That was last Monday.  On Wednesday I learned that the 'wired' versions were not being shipped yet so I called and left a message that the 'wireless' version would be fine

I haven't heard back yet!

The solar modules cannot be installed until I have the Tigo components, unless I decide to forego the monitoring, which I am loathe to do.  Nevertheless, if I cannot get cooperation from AEE Solar I may be forced to either go to a competitor (DC Power Systems) or live without the Tigo components.

Panel Optimization Part 2

My analysis of module optimizers is now complete.  If you recall, there were three under consideration:

• SolarMagic™ by National Semiconductor
• TIGO Energy™ Module Maximizer
• SolarMizer™ by Xandex

And the winner is...

Tigo Energy Module Maximizer-ES 60

It was actually a fairly straightforward decision.

All three optimizers provide power maximization at the module level. This will minimize the degradation of the power output for an entire series based on an issue for a single module, such as shading.

But the one feature that sets the Tigo product apart from the others is the module monitoring.  Much as the Enphase micro-inverters, the Tigo maximizers provide per-module information which is collected by a central sensor/combiner and managed by a Maximizer Management Unit (MMU).  Monitoring can be shared with Tigo so that their engineers can diagnose module problems should they occur.

Tigo components are approximately:

• $56 per maximizer.  One per module.
• $650 for the MMU.
• $0 for 6 months of monitoring by Tigo.
• $350 for 5 years of monitoring.

Module performance can be monitored using a web browser and the display can be customized so that individual solar panels are shown visually matching the actual installation.

Unfortunately, the technical information available from the Tigo website is minimal and the options are varied so decisions are tough to make without requesting additional information. For example, the website does not mention several important facts:

• A 'sensor/combiner' is required.  This is the box that communicates with each module maximizer and forwards that information for monitoring.
• The pictures give the impression that there are four connections to each but there is a fifth connection that is made to the 'sensor/combiner' over which the module control signals are sent and the module performance information is retrieved.
• The maximizers can communicate with the 'sensor/combiner' either wired or wirelessly.
• The interconnections between the modules is done via Tigo custom connector.  I'm sure that is deliberate so that incorrect connections cannot be made and damage the maximizer or the module.
• Nothing is said about how the 'sensor/maximizer' connects to the Internet for monitoring. Is a wired Cat-5 connection required or is WiFi supported? Can the monitoring information be collected locally or am I locked into Tigo's monitoring?

But the Tigo's are brand new and are only now becoming available so it's not surprising that information is only now trickling out and even dealers have little information. I've been working with Tigo directly and am now working one of their distributors, AEE Solar, to figure out the proper configuration. Once I get a configuration finalized I'll have to work with one of AEE Solar's local dealers to place the order.  

Panel Optimization

Perviously I mentioned that there are some minor shading challenges for the main house from a tall cottonwood tree.  That tree will cast some shade on the house on the non-summer days.  The Enphase micro-inverters would effectively address this problem but they are not available yet for the 210 watt panels that I'm specified.

An alternative technology has recently come on the scene in the form of 'module maximizers' (also called 'power optimizers').  Like the Enphase micro-inverter, one module maximizer is attached to each panel and performs "dynamic module balancing".

I know of three such products:

• SolarMagic™ by National Semiconductor
• TIGO Energy™ Module Maximizer
• SolarMizer™ by Xandex

The SolarMagic product purportedly reclaims up to 50% of lost energy due to shading.  TIGO claims "Up to 20% more Energy."  And Xandex says their SolarMizer can be selectively installed on only those panels subject to shading. More research is needed!

One immediate advantage of the TIGO system is the availability of individual panel system monitoring.  Monitoring is very important in diagnosing panel issues.  One large bird dropping can rob an entire series of panels half as effective.  Monitoring can quickly identify the panel in trouble so that you can remedy the problem.

The Enphase micro-inverters also offer monitoring. Monitoring is not mentioned for SolarMagic and SolarMizer.

I've been concerned about how to best monitor my system and have even considered building my own micro-controller board that talks to the inverter and sends data to my Macintosh once a minute (or so).  While that would be great fun it sure would be nice to have monitoring from Day One.

TIGO, like Enphase, offers a system monitoring device and web-hosting of data.  It costs: first six months free, then $350 for five years.  But it also offers remote system management including the ability to completely shut down the system for safety in case some service was required.

Another advantage of a monitoring system of this kind is that it is independent of the kind of inverter being used.

I'm off to do more research and the check availability.

Getting Serious

Since I'd decided to be my own solar contractor it was time to get serious. The primary objective was to make sure that I was not overlooking anything — those 'gotchas' that can really make a hash of things.

So I dusted off my copy of Photovoltaics: Design and Installation Manual and did all of the exercises pertinent to my project.  Photovoltaics is an excellent book and the only recommendations I could offer would be 1) add more checklists, and 2) eliminate the "Solar Data" section (just put the latter on-line).

Everything checks out okay:

• House:
• 24 panels (3 strings of 8)
• Medium size inverter capable of handling ~350 VDC
• Inverter to be placed immediately adjacent to the main power panel
• Garage
• 16 panels (2 strings of 8)
• Medium size inverter capable of handling ~350 VDC
• Inverter to be placed next to the auxiliary panel in the garage

About this time, after visiting with some of the neighbors, I started thinking about cutting back to just the house-mounted portion of the system and leaving the garage as a possible future expansion.  There would be no real penalties for postponing the garage and there were a few minor benefits: 1) I wouldn't have to get Architectural Committee approval for the project, and 2) I wouldn't have to remove a tree:

The garage is on the right and this picture is taken facing to the south-southwest.  The tree immediately to the left of the garage shades the south-facing roof of the garage too much and would have to be removed.

The house will have a bit of a tree-shading problem from a large cottonwood tree located in the southwest corner of our lot:

This picture is taken facing almost directly east.  The cottonwood can be seen beyond the garage and is about 60' tall and growing taller each year.  It will cast a shadow on a portion of the house roof during non-summer days until about 11 AM.  The Enphase micro-inverters would help minimize the impact of that shading but since they are not available in the size I need I will consider using Tigo Energy's Module Maximizer or National Semiconductor's Solar Magic Power Optimizer, both of which can be retrofitted after the initial installation.

Having decided to eliminate the garage portion of the project I'm ready to move ahead.  That means getting a final equipment bid for the panels, inverter and railings and talking to my electrician.

Professional Residential Solar Bid Analysis

I mentioned in an earlier post that I'd been disappointed by the bids I'd received from three of our local professional solar firms.  Now that I've gone over the various layouts and configurations it might be interesting to examine those bids and see how they stack up.

Bid #1 provided pretty good detail.  Bid #2 just provided a bottom-line number.  Bid #3 was essentially given by phone.  And 'My Est' was based on prices for components I could purchase as an individual (mostly from from Wholesale Solar) plus estimated labor by an electrician.  (The companies providing bids will not be identified.)

                      Bid #1    Bid #2    Bid #3    My Est
------------------  --------  --------  --------  --------
Total Wattage          5,250     4,800     4,800     8,400
Panels Total        $ 28,673                      $ 34,440
  Each                 $ 956                         $ 861
  Type            Solarworld     Sanyo       REC     Sanyo
  Wattage Ea.            175       200       200       210
  Count                   30        24        24        40
Inverters Total      $ 6,213                       $ 7,070
  Each                 $ 207                       $ 3,535
  Type               Enphase   Enphase   Enphase Sunny Boy
  Count                   30        24        24         2
Racking Total        $ 3,784                       $ 2,218
Labor Total          $ 2,870                      $ ~3,000
System Discount          $ 0       $ 0       $ 0  $ -3,269
------------------  --------  --------  --------  --------
Grant Total $       $ 41,541  $ 40,870  $ 38,040  $ 43,459
Cost per Watt         $ 7.91    $ 8.51    $ 7.93    $ 5.17

------------------  --------  --------  --------  --------
Federal Tax Credit  $ 12,462  $ 12,261  $ 11,412  $ 13,078
Net System Cost     $ 29,079  $ 28,609  $ 26,628  $ 30,381
Effective per Watt    $ 5.54    $ 5.96    $ 5.55    $ 3.62

For comparison purposes, at the time of writing this blog entry, the prices for various components from Wholesale Solar are as follows:

  Solarworld 175 W panels         $ 855
  Sanyo 200 W panels                N/A
  REC Solar 205 W panels          $ 564
  Enphase 175 W microinverters    $ 197
  Enphase 205 W microinverters      N/A

Doing the work myself (hiring out the electrical expertise and roof labor, of course) has a distinct advantage in terms of price per watt.

For only 4% more money I can get 60% more electrons!

It was pretty easy for me to make a decision: I'll be my own solar contractor and work with an electrician.

On Being a Neighbor

Looking over the house and garage configuration I realized that the detached garage installation might be something the neighbors must appreciate some conversation about.  Plus, our subdivision has CC&Rs* governing the type of roofing material we can use on our house.  Solar panels would certainly qualify a 'roofing material'.

Some counties and states have laws which nullify subdivision restrictions.  That may soon happen in Utah, but I would check with my neighbors even if there were no restrictions.

In order to moderate the CC&Rs our subdivision has an 'Architectural Committee', supposedly made up of a few neighbors with one servicing as the committee chairman.  In our case, the original subdivision developer is still the 'Architectural Committee' — a common situation.  (He tried to talk me into becoming the Committee!  Ha ha!)  So I called him and he was fine with my solar panels but asked me to canvas the neighbors and let him know their reaction.

I spoke with our most immediate neighbors and none of them offered any resistance to my solar plans though I noticed some slight concern by one.  So I decided to mull it over for a few days.

* — Codes, Covenants and Restrictions

Micro-inverters

Now that I had identified the best panel locations:

I carefully measured the available surfaces.  Then I started researching panel/inverter combinations with an eye toward which panels would conveniently fit in the available areas.  I'll spare you all the details of the research (many, many panels and all the various inverters) and simply say that I settled on the Sanyo panels because they seemed to offer the highest efficiency for a decent price with good reviews.

The Sanyo panels I originally looked at were rated at 205 watts.  Later, I settled on the 210 watt panels, known as HIT Power 210N or HIP-210NKHA5.  Each panel is 62.2" by 31.4" (1580mm by 798mm).  That's just about 1.26 square meters.  The individual cells are approximately 18.9% efficient and the entire panel hits 16.7% efficiency.

Playing around with various panel configurations resulted in the following two primary options for the house:

The 'portrait' orientation of the panels on the house would allow 26 panels, but the two panels closest to the left edge would become shaded a bit too early in the afternoon.  24 panels looks just right and would produce approximately 5,040 watts.  And the number '24' is also good for another reason we'll explore in a bit.

The 'landscape' orientation would allow a maximum of 27 panels but three of those panels would be marginal so we'll once again settle for 24 panels.  This orientation has a minor advantage of keeping the panels a bit further away from the left edge of the roof, postponing the effect of the shading of the adjacent roof line a bit.

Ultimately, I decided to go with the 'portrait' orientation primarily because of the mounting racks.  The rails upon which the panels are attached work best when they run perpendicular to the long dimension of the panel.  It is also easier to attach the rails to the roof trusses if the rails are aligned horizontally across the roof rather than up-and-down.

The garage configuration only works in the 'portrait' configuration:

The garage provides room for 16 panels, supplying about 3,360 watts.  '16' is a good number when combined with '24' from the house configuration.

The inverter decision has been somewhat difficult to make due to recent developments of the 'micro-inverter'.  Micro-inverters are small, single-panel inverters which convert the ~40 VDC output from one panel into 110 VAC.  Use of micro-inverters is considerably simpler than the use of a more traditional multi-panel inverter because one may simply combine the outputs of all of the small micro-converters and wire that directly into a breaker on a standard electrical panel.  If the utility power shuts down for any reason then each micro-inverter immediately shuts down providing safety for the utility and utility personnel.

Before the micro-inverter was invented the only choice was a more traditional inverter to which a series of solar panels are wired.  There is a wide variety of 'traditional' inverters on the market and careful calculations must be performed to match up the numbers of panels in a series and the number of series with the inverter.

In my case, 8-panel series works quite well.  The 24-panel house configuration will provide 3 series of 8 panels and the 16-panel garage configuration will provide 2 series of 8 panels.  It is critical for traditional inverters that each series be balanced, and the number '8' works well with '24' and '16'.

Carefully balancing the number of panels in each series would also be a great advantage in the future should I decide to switch over to a battery backed system.

I would like to use micro-inverters.  They are supplied by Enphase Energy.  One major advantage of the Enphase micro-inverter is that Enphase has a very nice monitoring system that allows you to see how each individual panel is performing.  This could save you a lot of time when trying to figure out why your system may not be performing as expected.

Unfortunately, the Enphase micro-inverters are not yet available for the 210 watt Sanyo HIT panels and it could be next year before they are shipping.  I'm not willing to wait and the price differential between the 205 and 210 watt panels is not compelling enough to step down to the lower wattage panel.

So I've decide to go with a traditional inverter.  More on the inverter later.

Panel Potentiality

The next step in my project was to identify where I could put PV panels.  Ground-mounted panels would only work in the front yard and I didn't think that the neighbors would be very happy with that solution.  Fortunately, there were several roof-mounted possibilities, on both the house and the small garage to the south of the house.

Each of the surfaces on the house is nicely oriented to the south, and that on the detached garage is only a bit off of south.  The ideal orientation is directly south so all four of the marked potential sites are a win in that regard.

A second win came in the angle afforded by the roof surfaces.  The ideal inclination of solar panels is the angle of latitude of the site.  Our latitude is 39.5° and the roof surface is at an angle of 40°.  It could hardly be better!

'A' is by far the best location and offers several advantages:

• No shading from trees.
• Presents a large contiguous space reducing wiring requirements.
• Can support 8 to 10 kW.
• Very convenient to the service panel.
• Very easy to access for cleaning and maintenance.

Unfortunately, 'A' is also exposed to the street and would distress the neighbors.  We have an architectural council and CC&Rs in our subdivision and I was pretty sure it would be difficult to get approval for this area.  I might consider 'A' for a future expansion should the laws be changed to bypass local restrictions but I would still be reluctant to do something that would cause friction with the neighbors. Eliminated.

'B' is a nice location, too.  There is a minor shading issue from some tall trees to the east in the early morning, and from the roof line to the west in the evening.  The tree shading will be minimal during the winter.  Though not as convenient to the service panel, it's still workable.  And being on the back of the house eliminates any neighbor concern.  It can support about 5 kW and is also very accessible for cleaning and maintenance. To be considered.

'C' is not a nice location.  It's a long distance from the service panel, supports only about 2 kW, and is not accessible for cleaning and maintenance.  Eliminated.

'D' is a decent location, easily accessed for cleaning and maintenance, convenient to a service panel in the garage, and nicely rectangular.  There is a tree just to the south-east which would need to be removed in order to reduce shading.  It can support about 2.5 kW of panels.  Unfortunately, it's exposed enough to the street that I would not feel comfortable committing to this site without first talking with the neighbors.  To be considered.

'B' and 'D' together could provide about 7.5 kW which would chop the most expensive half of my annual usage under average weather conditions.

So 'B' and 'D' are under consideration.

Solar Education

Determined now to figure out what it would take to tap into the sun's power for residential electrical generation.  What I discovered is that it is no longer 'rocket science' and that there are commercially available solutions.  I also learned that it's still an immature industry and that there are only a handful of installers in our area.  And finally, I learned that the Federal Government's 30% renewable energy tax credit is indeed ramping up interest and causing some interesting responses by the installers.

Not Rocket Science

• Everything is basic electronics.
• Solar panels
• Inverters
• Wiring
• Safety components
• High-voltage DC is usually involved.
• The concepts are relatively easy to understand for someone with an engineering background.

Commercial Solutions

• I would only ever use commercially available, off-the-shelf components.
• Solar panels come in a variety of types, configurations, prices and efficiencies.  They all output DC.
• There are many types of inverters available.  Some connect to your utility and some charge up batteries.  (I will only discuss the former, also known as grid-tied inverters.)

Installers

• People are anxious to get into the solar installation business.  There's good money to be made, particularly from those who are proud 'early adopters'.
• The three installers in my area from whom I got a bid unanimously marked up the system components by 30% (in addition to specifying a labor fee).
• There are some important concepts to understand before embarking on a solar project and these installers generally understand those concepts.

Where can you go to learn about residential solar?

• Subscribe to Home Power Magazine.  Get the on-line subscription and read the archived articles.
• Purchase Photovoltaics: Design and Installation Manual from SEI (Solar Energy International).  Make sure to get the most recent edition.  Do the exercises.
• Now go back and re-read some more articles in Home Power Magazine.
• See if your local community college has any courses on solar power.
• Talk to people who have already invested in solar power.
• Rinse and repeat.

I'm an engineer and have fooled around with electrical systems since I was a kid and I'm comfortable doing most of the work associated with this solar project, but that's only after studying and researching for about two years.  Even with that training, however, I'm not a licensed electrician and so will hire and supervise an electrician for the installation.  It's unlikely the electrician with whom I work will have any solar PV system experience nor will he have any high-voltage DC experience.  But the principles are straightforward and after reviewing the code and NEC requirements I am confident in a safe and reliable outcome.

If you are thinking about doing a solar PV system on your own then you certainly should be very comfortable with:

• Solar panel and inverter sizing calculations.
• 110/220 VAC systems.
• Interconnecting with the utility.
• Handling high-voltage DC (~400-600 VDC).
• Truss load bearing for a roof-mounted system.
• Local code and permit requirements.
• Utility interconnect requirements.
• Subdivision architectural/CCR restrictions.

Then Google Happened

I joined Google as a software engineer in early 2007.  One of the most enjoyable benefits (in addition to the fabulous food) is the vast array of 'Tech Talks' presented by Googlers and their friends.  There were a series of Tech Talks dealing with taking advantage of solar energy, including solar PV systems, which I attended.

One such Tech Talk, presented by a couple of Googlers, discussed their own residential solar PV investment and covered the cost, the effects of shading, equipment, and much more.  Google itself had started an installation of solar PV at their Mountain View facility which would ultimately have the capacity to produce up to 14.5 mW of electricity.

Another Tech Talk presented new technologies being developed that looked very exciting.  One particular technology in which Google has invested uses a printing technique for making sheets of PV material with a per-watt cost approaching $1.

Cost is one critical factor when deciding if a solar PV investment is reasonable.  Historically, PV panels have been very expensive, affordable only by NASA and a few remote facilities where shipping fuel in would be impossible.  Current prices for panels are hovering around $4 per watt.  But wait, there's a lot more involved:

• Inverters
• Protective devices
• Racking
• Batteries
• Synchronization devices
• Cabling
• Installation

Add all of this up and the final per watt cost will range from $7 to $13 per watt.

At those prices you can estimate that a 5 kW system would cost from $35,000 to $65,000.

Not cheap, eh!

Enter Uncle Sam with their 30% tax credit for renewable energy systems.  Get it while you can!  That would reduce the costs of a 5 kW system to something in the range of $25,000 to $46,000.

I'll do a little cost-benefit analysis in a future installment.

Early Ideas

"There's just got to be a way to tap into the energy of the sun!"  This idea drove me to consider many outlandish ideas.  How about a 'solar trough' — a long, parabolic reflector with a steam pipe at it's focal point?  This design would simplify the tracking while maximizing the captured energy.  I even designed a simple tracker with only two photodetectors.  Heat the liquid in the pipe, drive a turbine, and voilá!  Electricity!

If only it were so simple.

So back to more traditional approaches, like passive solar collectors for hot water and solar PV panels for electricity.

I've ignored the passive hot water system for now primarily because we just don't use that much hot water, natural gas is still cheap, and installation would not be trivial.

At this point in my life (about 12 years ago), solar PV systems were still pretty much in the 'emerging' category of development.  And while I don't mind being on the 'bleading' edge of technology, there were some significant challenges to overcome:

• Grid-tied systems were in their infancy.
• A battery-based system would not meet our total needs.
• The cost was prohibitive.
• PV efficiencies still left much to be desired.

And there was plenty else to keep me busy.  So I took a "let's wait and see what happens in the technology" attitude.

Introduction

Tapping the energy of the sun has been of great interest to me since I was a young boy burning up paper with my first magnifying glass.  Technology has come a long way since those days and is now such that the proposition of tapping into that big yellow disc in the daytime sky is within reach, perhaps with a bit of stretching, but possible nonetheless.

Google has inspired me!  I work for Google and am proud of Google's investment in Solar PV (photovoltaic) electricity production.  After attending several Tech Talks and watching numerous videos about residential solar I decided to get serious about my own system.

I don't have a system yet—I'm in the middle of the project.  So it's a great time to start recording my journey.

This blog will first go back in time and describe the path of my awakening to the solar potential.  Then it will bring us up-to-date with my investigations into different solar solutions, technologies, battery-based and/or grid-tied, utility interconnect, city and state hurdles, installers, tax credits, layout, and so much more.

So welcome and I hope my experiences will help you as you consider residential solar of your own.