SUN GETS EVEN

Break-Even Cost for Residential Photovoltaics in the United States: Key Drivers and Sensitivities
Paul Denholm, Robert M. Margolis, Sean Ong, and Billy Roberts, December 2009 (National Renewable Energy Laboratory)

SUMMARY
One of the truest observations ever made about solar energy (famously attributed to Nathan Lewis of the California Institute of Technology (Cal Tech) by NY Times columnist Thomas Friedman) is that, unlike the emergence of the hand-sized cellular phone which was such a breakthrough people were immediately willing to pay whatever it cost, solar energy-generated electricity does nothing "new." People already have electricity. Why should they pay more for it?

Concerns about global climate change and other degradations associated with fossil fuel-generated electricity somewhat diminish the profundity of Professor Lewis’ observation. But only somewhat, which is why researchers at Cal Tech and many other innovation centers around the nation and around the world are working hard to achieve a grid parity cost for solar energy.

The grid parity point (also called the break-even cost) is defined in Break-Even Cost for Residential Photovoltaics in the United States: Key Drivers and Sensitivities, from the U.S. Department of Energy (DOE) National Renewable Energy Laboratory, as the point at which the cost of solar energy-generated electricity from residential photovoltaic (PV) panels is even with the cost of electricity purchased from the grid.

click to enlarge

It turns out residential PV electricity has achieved grid parity in some places and under some circumstances. But in other places and under other circumstances, it is far from parity. The variables influencing the prices of electricity and PV electricity are so wide-ranging that the grid parity point can be as low as $1 per watt or as high as $10 per watt.

The NREL researchers looked at the price of electricity in late 2008/early 2009 at the biggest 1000 U.S. utilities to calculate this wide variation in grid parity points and noted the solar resource across the country does not vary by nearly so much. They found that the key drivers of the grid parity point for PV-generated electricity are not technical factors (solar resource, rooftop orientation) but non-technical considerations such as the cost of electricity from the grid, the rate structure, or the cost and availability of PV system financing.

The NREL paper includes a hypothetical assessment of what grid parity might look like in 2015 when there are more efficient PV systems and policies that put a price on greenhouse gas emissions.

click to enlarge

COMMENTARY
The NREL study first established a base-case concept of grid parity by incorporating the 1000 biggest U.S. utilities’ late 2008/early 2009 rates (95% of the total residential load) to determine a grid parity cost for PV-generated electricity in dollars per watt ($/W) and the grid parity electricity price in cents per kilowatt-hour (cents/kWh). The researchers ran cases of variables in which the net present cost (NPC) of the PV system equaled the net present benefit (NPB) to the system’s owner in both $/W and cents/kWh.

The cost calculations considered everything that went into what was paid for the system, including financing and incentives. The benefits were the reductions in the electricity bill.

The cost was based on a PV solar system financed with a 20% down, 30-year home-equity type of loan with a tax-deductible interest rate of 5% and a 28% marginal federal tax rate. It also assumed the current 30% federal investment tax credit (ITC) was in place, along with known state, local, and utility incentives and standard accounting procedures.

The benefit calculations were based on the 2008 installed cost of $8/W (and an inverter replacement at 10 and 20 years) with a south-facing system, panels tilted at 25 degrees, and a 0.5% per year degradation. This would lead to a progressive reduction in the owner’s electricity bill over the life of the system dependent on the system’s performance and the local utility’s electricity price.

click to enlarge

System performance evaluation began with the consideration of regional insolation. This was converted to solar energy production and an average system factor of 77%.

The benefit was calculated for a flat, seasonally-adjusted electricity rate and for a time-of-use (TOU) rate structure and incorporated an “average” cost of electricity to residential customers by an evaluation of a range of utilities’ prices, state by state.

At the point in time when the study was done for the base case, with a flat electricity rate, only ~11% of residential electricity sales - from utilities in areas of the country with high electricity prices and good solar resources (California) or high electricity prices and strong incentives (New York, Massachusetts) - were at or near grid parity if the system cost is $8 per watt. If the system cost drops to $6 per watt, ~42% of residential electricity sales - from utilities in areas (Florida, North Carolina, New Jersey, a growing part of the southwest) - were at or near grid parity.

Only the part of the customers of those utilities with all the criteria (full retail net metering, good solar exposure, and assumed financing) would be at grid parity. Up-front costs further limit the customers in the grid parity group and caps on incentives would continue to limit that number.

click to enlarge

Adding a time-of-use (TOU) rate structure to the calculation complicated it. Only about half of the states have TOU pricing. Some TOU structures do not reduce the prices paid for electricity and may even add to the benefits of a solar system for some customers, such as those who need a lot of electricity during the day.

With TOU, grid parity for an $8/W system got to ~19% of electricity sales and ~45% for a $6/W system, mostly in the same Southwest and Northeast states plus a few others (Wisconsin, Florida).

The NREL researchers studied other variables, including higher electricity prices and lower system costs. Both of these are realistic future possibilities. The percentage of electricity sales went up under either circumstance, This explains why the PV industry so wants national policy measures such as (1) a Renewable Electricity Standard (RES) (that would drive volume growth, create economies of scale and reduce system costs) and a climate bill (that would cap greenhouse gas (GhG) emissions and make electricity generated by GhG-emitting sources more expensive).

click to enlarge

The researchers’ hypothetical 2015 scenario assumed a future set of assumptions:
(1) The price of electricity rises 0.5% per year and is 3% higher in 2015.
(2) No state incentive programs for solar PV systems are available.
(3) Emissions cost $25 per ton of CO2e, adding 0.3 cents/kWh (Oregon) to 2.5 cents/kWh (North Dakota) for a national average of 1.5 cents/kWh.
(4) Solar system efficiency is improved along presently expected lines so that the maximum cost is $6/W.

Under these hypothetical 2015 circumstances, ~43% of residential electricity sales came at grid parity for areas at $4/W system costs and ~85% of sales came for areas at $3/W system costs.

TOU rates in 2015 would bring significantly more electricity sales to the grid parity level, 75% at $4/W system costs and 91% at $3/W system costs, because it was assumed all states would be using TOU pricing by then.

click to enlarge

DOE’s target installed cost for 2015 is $3.50/W. Assuming that is achieved, there is likely to be grid parity for some electricity customers in most of the Southwest and New England (~67% of all U.S. residential electricity sales) if the 30% federal ITC remains in place (as it is scheduled to) and if there is a $25 per ton price on GhGs. Adding in TOU rates would bring that to ~88% of sales.

The NREL paper includes a consideration of general ways (a) technical performance, (b) electricity cost, (c) financing and (d) policies might vary going forward and alter these conclusions.

click to enlarge

General observations from the paper’s conclusion:
(1) Grid parity conditions will emerge first in the Southwest, driven by resource, and in the Northeast, driven by high electricity prices.
(2) As PV system cost declines, grid parity will emerge in the Southeast and Midwest.
(3) Very low electricity prices will likely prevent grid parity in some Northwest and Midwest areas despite a system cost of $3.50 per watt and the federal ITC.
(4) The calculations do not account for the potential of a “deep, sustained market.”
(5) Even grid parity will not necessarily prevent high up-front costs from discouraging widespread PV adoption.
(6) Better financing options could alter adoption.
(7) Further analysis that includes (a) a “demand curve” for PV at various price points and (b) the commercial buildings rooftop PV market potential must be done before complete conclusions about PV grid parity can be made.

click to enlarge

QUOTES
- From the report’s conclusion: “…the current break-even [grid parity] price varies by more than a factor of 10 even though the solar resource varies by less than a factor of two. This difference is largely driven by incentives, which can exceed $5/W, and the difference in electricity prices, which can vary by a factor of eight (or more when considering the range of tiered rates in California). Even without incentives, large variations in break-even [grid parity] cost will remain given the range of financing options and other non-technical factors.

click to enlarge

- From the report, regarding its 2015 scenarios: “The PV market of the future will have a variety of customers with different financing options, homes with non-optimal orientations, and changes in electricity prices and rates. Many of these drivers, such as escalation of electricity prices and carbon policies, are highly uncertain. As a result, it is important to consider the sensitivity of the break-even [grid parity] price to a variety of drivers.”

click to enlarge

- From the report’s conclusion: “…the scenarios do not consider the potential for a deep, sustained market. Therefore, caution must be used when considering this analysis. PV breakeven [grid parity] does not imply that customers will necessarily adopt PV, and only a fraction of customers in each utility will have the necessary combination of good solar access and attractive financing options. A true depth of market analysis is required to determine a “demand curve” for PV at various price points. This must be combined with analysis of commercial buildings to provide an estimate for the market potential of rooftop PV.”