- First Solar, a long-time manufacturer of cadmium-telluride photovoltaic (CdTe PV) modules, has dramatically overhauled its main business model in recent years – all in response to simple energy economics.
- Solar power generation has basked in the balmy weather of government sponsorship, a low-interest rate capital market, a high-tech reputation, and environmental popularity for the past decade or so.
- While offering the seeming advantages of no fuel, waste, or repair costs with a distributable power base for 'green' electricity, solar economics could stand closer scrutiny.
First Solar's principal product, the CdTe PV module, measures two feet by four feet in size, generating 91 watts at a cost of $0.63 per watt - asserted to be a low-cost solar converter. Wattage output (Pending:W) for the device has been improving about 7% each year. Since FSLR has been guaranteeing at least 90% installation performance, let us conveniently consider this eight-sq.ft. mechanism as producing 80W, or 10W per sq.ft., setting the economic stage.
Earmarking 10am to 2pm as 'peak-generation' time, we observe a four-hour day, a 30-hour week, and a 1500-hour year of power generation at the 10W / sq.ft. rate. Annual sq.ft. energy conversion thus totals 15 kW-hours which can be sold at $0.10 per kWh, meaning that one square foot of PV cells can bill revenues of a paltry $1.50 per year! It seems reasonable that a square foot of farmland could yield a least a $2 bag of peanuts or potatoes every harvest.
Consider a further comparison. The W.H. Sammis coal-fired plant in southeast Ohio powers the grid at a 2000 MW rate, 24 hours a day, rain or shine, cloudy or not. This full-time baseload provider runs 168 hours a week, 8000 hours yearly, on a property less than one square mile. Conveniently apply the 2000 MW over 25 million sq.ft., and we have a 'coal-fired module' of 80W per sq.ft. - eight times more spatially efficient than CdTe PV's 10W per sq.ft.!
Now let's include the time factor: Sammis always runs - five or six times as much each year! Annual square-foot energy production surpasses 600 kWh, billing yearly revenues in excess of $60 - a revenue productivity advantage of 40 times!
Solar backers may counter: "What about the profits? Solar has no repair, waste, or fuel costs." Fair enough. Many utilities earn a 10% pre-tax margin or $6 annually per sq.ft. The CdTe PV module costs $0.63 / W or $6.30 / sq.ft. depreciated over 20 years - deduct $0.30 or 20%. Annual square-foot profits? First Solar's module is $1.20; Sammis is $6 - five times greater!
A crucial investment question arises: Can the PV module greatly boost its productivity soon?
Run time: FSLR recently acquired a solar-tracking manufacturer. Re-aiming the PV-cell arrays to trace solar trajectories (azimuth and altitude angles) can boost output by about 25%.
That now leaves 1/4th the productivity. When may PV-module outputs quadruple to match?
Energy density: First Solar's CdTe PV module wattage has been advancing five or six watts per year - a growth rate of 6 or 7% per annum. Doubling the output - barring a major breakthrough - will take ten years. A matching four-fold improvement is 20 years away.
With such an extended period of uncertain economic viability looming, investment returns in solar electric power seem quite limited and very risky. Could FSLR be the 'survivor' during the next two decades? FSLR may very well have the technology and finances to survive. However, we will leave that for some other analyst to investigate. As much as we admire leading companies, our investment principles steer us clear of difficult businesses like solar.
Additional evidence of solar industry 'headwinds' includes:
The cost of money in capital markets - the interest rate - is more likely to rise than fall.
Unsustainable sovereign debt levels may curtail government solar subsidies going forward.
Fossil-fuel prices may well decline from the 'casino-like' futures-market hyperinflated levels.
PV cells may have their own environmental issue: cadmium-telluride is hazardous material.
FSLR itself is 'battening the hatches to weather the storm':
The firm has restructured its business model from module sales to system-grid servicing. There is too much module inventory chasing too few solar power customers and installers.
FSLR closed its German production line: Europe lacks the extra land and sunshine, does not need much air conditioning, and they like their nuclear plants - 'ditto' for S. Korea and Japan.
Connecting to the power grid and managing electrical loads with cyclical sunshine is not easy.
FSLR recently issued shares to raise $400 million. We prefer 'earnings-financed' operations.
As a final observation, solar PV modules may have a role as auxiliary 'peaking' units to help fossil-fuel facilities manage the summertime air-conditioning loads. In such a case, the solar supply aligns with demand. A favorite utility - MidAmerican Electric, a subsidiary of Buffett's Berkshire Hathaway (BRK.B, BRK.A) - is a 10+% customer of First Solar. Unlike most utilities, MidAmerican retains all its earnings for expansion, and MAE's bright executives wield a very sizable wallet. If MidAmerican ever boosts its business with FSLR, this outfit may be worth a second look. MidAmerican understands energy-value propositions as well as anybody. As of 2014, no additional MidAmerican solar projects with FSLR appear to be in the pipeline.
Epilogue: To duplicate the Sammis plant's 2000 MW continuous capacity, there must be a 40 square-mile layout with some huge battery storing the energy for nighttime, rainy-day, wintertime distribution. Until that transpires, solar energy may logically be considered to be - for better or worse - a 'land-consuming, tropical-tanning, air-conditioning' sort of enterprise.