Due to the state’s revolutionary 2019 Building Energy Code requiring solar power to be installed at time of construction for all newly built homes, over time California home buyers could see per unit prices drop to rates approaching that of some of the world’s largest utility scale solar installations ($1/W in the United States). And if this seems hyperbolic, first let’s lay out the case for why this move by California is about to revolutionize residential solar in our country.
Outside of energy-nerd circles (used here endearingly of course), the dramatic and continuous fall in the price of solar panels is beginning to be acknowledged by the general public. Known as “Swanson’s Law” (a cousin of sorts to the more famous “Moore’s Law”), since the mid 1970s solar panels have gone down about 20% in price every time the volume we ship annually doubles. Put more simply, solar panels are at least three times cheaper than they were just a decade ago. While remarkable, and with every indication that this astounding trend will continue (some data shows the decline speeding up), that’s still just the panels.
What’s different here in the Golden State can be described in just two words: soft costs. For rooftop solar, some of these soft costs include permitting, financing, installation labor, cost of customer acquisition, paying suppliers, … and then of course hopefully some profit is left over for the company if they want to stay in business. Because these soft costs are greater for smaller installations, the U.S. Department of Energy (DOE) estimates soft costs can be almost two thirds of total costs.
Furthermore, soft costs aren’t decreasing as fast as hardware costs. However, this move by California is likely to bring down soft costs sharply, as it is expected to put downward pressure on nearly every one of these soft cost categories. For newly built homes, in many cases it eliminates them altogether.
The first piece of evidence comes from data provided to us by DOE’s National Renewable Energy Lab (NREL) in their annual U.S. Solar Photovoltaic System Cost Benchmark: Q1 2017.
We’re going to use this document as our base of discussion because the $3.17/W value that the California Energy Commission (CEC) decided (see page 38) to use in describing the potential costs per home for solar power, was based on the 2016 version of this document (including 2% inflation from that 2016 value).
Our standard solar power system is 3,015 watt system, composed of nine 335 W panels. It is assumed this power class will represent the base Tier 1 product available to the market. We chose this value as it is close (slightly larger) to our estimates of what the CEC is requiring under the new code.
pv magazine chose to use values for net profit, overhead and sales & marketing for the new home solar that match up with the standards per the National Association of Home Builders cost to build a home in 2017. And it is here that we see the fat fall away. The residential retrofit solar power industry simply cannot compete with the efficiency of including an extra feature in an already being built new home. This is not an attack on the creative engineers that figure out these solutions – but a recognition of reality. Savings: 81¢/W!
Permitting/Inspecting/Interconnection costs – instead of costing 10¢/W and requiring whole departments, this couple of sheets of paper will be piggybacked on an already existing paperwork process, including an interconnection application. These costs are estimated at 2% of all hardware and labor costs. Savings: 8¢/W.
Install labor costs will halve due to the efficiency of building at time of construction. You can almost hear the sigh of relief from many solar installers, not having to snake their body through the sweltering, insulation filled attic with your skin inches away from the shingle penny nails that poke through exposed on the interior side of the attic crawl space. Since California has higher labor costs than the national average, we used labor costs of $25.09/hr and electrician costs of $38.22/hr, as per NREL. Labor time was estimated at four hours, electricians at three, plus one more for commissioning. An hour for designing the layout was also included. Savings: 19¢/W.
Supply chain costs were looked at closely by NREL when they decided 43¢/W as the standard value. Local installers, buying from local distributors pay multiple layers of markups as products move through many layers. National homebuyers, reaping the benefits of economies of scale much like utility scale developers, will pay a standard 15% markup for shipping and storage. This 15% value is noted as the base value by the NREL report. It is a valid argument that smaller and custom builders will pay some of these fees. Savings: 34¢/W.
Electrical balance of system costs are a large variable on every project because homes are so different, and the upgrades needed vary so greatly. Will we need a new electrical panel? How will we snake the conduit from the roof to the inverter? Is the house up to code? In a new house, all of this is clean and the panels can be planned into the overall construction plan of the house. In fact – the basic costs here are as is: 30 feet of conduit and copper, a single external disconnect switch, a 20 amp fuse and $100 worth light goods (boxes, washers, bolts) to connect the wires. Savings: 18¢/W.
Structural balance of system is really just racking. And 11¢/W for flush mount residential racking is a great price – but even that is expensive compared to what can be done with a clean canvas.
The above QuickmountPV system needs a total of sixteen of their Quick Rackassembly units. You need nothing else but to attach your solar panels afterwards. These units can be bought in bulk, direct from manufacturer, probably at $10/each by a national home buyer. Savings: 6¢/W.
The NREL report suggests standard inverters can be purchased at 13¢/W – and that’s what we went with in this report. There is an argument that module-level electronics will be needed in order meet Rapid Shutdown requirements in NEC 2017 – this could be met on our model project at 4-8¢/W with various solutions. Savings: 6¢/W.
These numbers are not far from what Andrew Birch, former Sungevity CEO, showed us in Australia recently – $1.34/W to retrofit residential solar power versus $2.90 to do the same work in the California, mostly due to savings from those same soft costs that U.S. home builders have obviously figured out.
And there still are a few black swans out there that could drop pricing further. Will an integrated solar+housing material manufacturer like Tesla or RGS Energydeliver a product that combines the cost of racking+structural BOS (40.3¢/W), while also dropping solar labor installation costs (3¢/W) by combining that with the roofing material install time? Tesla’s Solar Roof does this, however, as of yet it is aimed at premium roof replacements and will seemingly cost much more than the base installs with commodity hardware modeled here.
Adding it all up we come to one last number – the California Energy Commission projects* that a 3,015 watt solar power system will generate, roughly, 133,630 kWh in a 30 year lifetime (4,785kWh/year 1, 0.5% annual degradation). At a system cost of $3,381 – that’s a lifetime cost of electricity at 2.5¢/kWh.
The CEC noted that systems at $3/W will cost new homeowners about $40 a month in a 30 year mortgage. Using these calculations the monthly mortgage payment will be $17 versus the same $80/month savings projected.
*The output of pv systems varies depending on climate and location within California, per the charts starting on page 39. For purposes of this document, we averaged the statewide numbers and then scaled them to our 3kW system.
These values are before incentives. In 2020, the homeowner would be able to claim a 26% tax credit against their solar system costs and in 2021 22%. After that, the residential solar tax credit expires. There are no state level solar power incentives in California – just expensive electricity and time of use rates.
Criticisms of the mandate
The biggest criticisms of California’s requirement seem to largely fall into two camps. Some academics, intellectuals, and other garden variety wonks have decried this move because it doesn’t align perfectly with their espoused pet framing. These are largely bereft of any real salient or tractable points, and come off as sour grapes because the naysayer wasn’t first consulted for his self-aggrandized oratory.
From the other flank comes the sophomoric dollar to dollar comparison of constructions costs for rooftop solar versus utility scale. These arguments are championed by those with a requiem for a vision of wide open swaths of landscape striped with clean lines of panels on aluminum rack-mounting laid out with mathematical purity like a giant monocrystalline Zen garden that’s been freshly raked.
Looking at the contrasting hot mess of residential one-off installations, a literal piecemeal, it is easy to see why the proponents of this utility-scale-is-the-best-scale position immediately throw out a dollar-to-dollar comparison with the current cost norms of residential solar construction. This ignores the fact that residential solar generates electricity directly at the point of use (the sun already has an innate distribution system) while utility scale is still dependent on transmission and distribution from the grid. The US grid is estimated to be a $5 Trillion dollar asset, and utility-scale solar and other forms of centralized generation get that for free in the calculation (not to mention all other non-generation costs).
Remember, distributed energy and efficiency just saved a $2.6 billion upgrade to that $5 trillion grid asset in California.
The only lucid comparison for residential solar prices are not the wholesale cost of utility-scale generation, but the retail rates of electricity paid by the consumer. This requires a necessary shift in thinking. The default perspective for most energy wonks is to first consider the generation asset, and then the journey of the produced electricity through the grid, and then finally almost as an afterthought we note the consumed energy service. Instead we should start with the point of use for the consumer’s energy service need, then we trace backwards to see how that need can best be met.
This monumental move by California shifts the playing field in unprecedented ways. Now residential solar will be much better able to leverage economies of scale across a wide variety of soft costs in ways that just weren’t possible before. Even for the wonkiest of wonks this will likely require a fundamental shift in thinking about how we characterize these assets and energy services.