8 The Circuitous History of Photovoltaics
“The solar industry did not follow a straight path from laboratory to commercial success. It took wrong turns, lost decades, and depended on policy support that was often inconsistent, sometimes irrational, and ultimately transformative.”
The physics of photovoltaics was understood by 1954. The technology was “ready” by the 1970s. Yet solar remained expensive and marginal for another 40 years. The gap between technical possibility and commercial reality—and how it was eventually closed—reveals the complex interplay of policy, markets, and manufacturing scale that our Framework for Change emphasizes.
8.1 The Space Age: 1958-1973
8.1.1 Vanguard 1
The first practical application of solar cells came just four years after the Bell Labs breakthrough. Vanguard 1, launched in March 1958, carried a 0.1-watt solar cell array as backup power. When the primary batteries died after a few weeks, the solar cells continued operating—for six years.
For space applications, solar cells were uniquely suited:
- No atmosphere to block sunlight
- No alternative (nuclear too heavy, batteries too short-lived)
- Weight, not cost, was the limiting factor
Space agencies and defense contractors drove early development. Cells improved from 6% efficiency (1954) to 14% by 1970. The market was tiny—a few megawatts total—but prices were irrelevant when the alternative was mission failure.
This initial phase illustrates an important pattern: early technology development often occurs in niche markets where the technology’s unique advantages outweigh its cost disadvantages. Space provided a protected market that funded basic R&D, efficiency improvements, and manufacturing learning.
8.1.2 Oil Embargo and the Solar Dream
The 1973 oil embargo catalyzed interest in terrestrial solar. Suddenly, energy independence became a political priority. The U.S. government launched research programs, built demonstration projects, and imagined a solar future.
President Carter installed solar water heating panels on the White House in 1979 (President Reagan removed them in 1986). The Solar Energy Research Institute (now NREL) was established. Photovoltaic costs fell from $100/W in 1975 to $20/W by 1980—impressive progress, but still 100× too expensive for grid competition.
Then oil prices collapsed in 1986. The crisis mentality evaporated. Federal support dried up. Solar entered what some call the “lost decade”—the 1980s and early 1990s, when annual global production stagnated at 50-60 MW.
Global PV production in 1985: ~50 MW At $10/W module price: $500 million market Contrast with global oil market: ~60 million barrels/day × $20/barrel × 365 = $440 billion/year
Solar was 0.1% of fossil fuels’ market size. No wonder it attracted little attention when oil was cheap.
8.2 Japan: The Rooftop Program (1994-2004)
8.2.1 Subsidies That Worked
Japan, lacking domestic fossil fuel resources and scarred by oil shocks, maintained solar support through the cheap-oil 1980s. But the breakthrough came in 1994 with the introduction of the Residential PV System Dissemination Program.
The program offered:
- Direct subsidies covering 50% of system cost initially
- Declining subsidy rates as costs fell
- Net metering allowing homeowners to sell excess power
The results were striking:
| Year | Cumulative capacity | Subsidy per watt |
|---|---|---|
| 1994 | 0.03 GW | $3.50 |
| 1997 | 0.1 GW | $2.80 |
| 2000 | 0.3 GW | $1.50 |
| 2003 | 0.9 GW | $0.40 |
| 2005 | 1.4 GW | Program ended |
Several design elements proved critical:
Degression: Subsidies declined annually, creating urgency and rewarding early adopters while avoiding perpetual dependency.
Learning-by-doing: Every installation provided feedback on installation practices, permitting, and grid integration. Japanese installers became skilled; Japanese manufacturers achieved scale.
Market creation: Stable demand enabled manufacturers to invest in production capacity, driving down the learning curve.
By 2004, Japan had more installed solar capacity than the rest of the world combined. Japanese companies—Sharp, Kyocera, Sanyo—led global production. Japan had demonstrated that policy could bootstrap a market.
8.3 Germany: The Feed-in Tariff Revolution (2000-2012)
8.3.1 The EEG: A Policy Innovation
Germany built on Japanese experience but pushed further. The Erneuerbare-Energien-Gesetz (EEG, Renewable Energy Sources Act) of 2000 introduced the modern feed-in tariff (FIT):
- Guaranteed prices for renewable electricity for 20 years
- Prices set above market rates to ensure profitability
- Grid operators required to purchase all renewable power
- Degression: rates declined ~5% annually for new installations
The initial solar FIT was approximately €0.50/kWh—vastly above wholesale electricity prices of €0.03-0.05/kWh. Critics called it absurdly expensive. Proponents called it investment in the future.
8.3.2 The Deployment Explosion
Germany’s solar deployment exceeded all expectations:
| Year | Annual installation | Cumulative capacity | FIT rate (€/kWh) |
|---|---|---|---|
| 2000 | 0.04 GW | 0.1 GW | 0.51 |
| 2004 | 0.7 GW | 1.1 GW | 0.57 |
| 2008 | 2.0 GW | 6.1 GW | 0.47 |
| 2010 | 7.4 GW | 17.3 GW | 0.39 |
| 2012 | 7.6 GW | 32.6 GW | 0.19 |
By 2012, Germany had installed more solar capacity than the entire world had in 2008. A cloudy country at 50°N latitude had become the world’s largest solar market.
The costs were substantial. German electricity consumers paid a “EEG surcharge” that reached €0.06/kWh by 2015—about 20% of retail electricity bills. Total cumulative cost of solar subsidies: over €100 billion.
8.3.3 Was It Worth It?
The German FIT debate continues today. Critics argue:
- Germany paid far more than necessary
- Early FIT rates were too high for too long
- The subsidies primarily benefited Chinese manufacturers
- The same money could have reduced more emissions elsewhere
Defenders respond:
- Germany bought down the global learning curve
- The subsidy per ton of CO2 avoided declined dramatically
- Other countries benefited from German-funded cost reductions
- No other mechanism would have achieved scale so quickly
The FIT surcharge was regressive: all electricity consumers paid equally per kWh, but wealthy homeowners disproportionately benefited from installing solar and receiving high FIT payments.
A family renting an apartment paid the surcharge but couldn’t install solar. A wealthy homeowner received €0.50/kWh for power worth €0.05/kWh—a massive wealth transfer from non-owners to owners.
This equity dimension fueled political backlash. The 2012-2014 reforms sharply reduced FIT rates and introduced caps on deployment. Germany’s solar market collapsed from 7.6 GW/year (2012) to 1.9 GW/year (2014).
The lesson: effective climate policy must address distribution. Policies that ignore equity create political constituencies for reversal.
8.4 China: Manufacturing Dominance (2008-Present)
8.4.1 The Strategic Shift
While Germany created demand, China captured supply. Chinese solar manufacturing grew from negligible (2005) to dominant (2015) through deliberate industrial policy:
State support: Cheap land, subsidized electricity, low-interest loans from state-owned banks. One estimate: Chinese solar manufacturers received $47 billion in state support between 2010-2018.
Integrated supply chains: Polysilicon production, wafer manufacturing, cell processing, and module assembly clustered in Jiangsu and Zhejiang provinces, minimizing logistics costs.
Aggressive scaling: Chinese factories were built at unprecedented scale, achieving cost advantages unavailable to smaller competitors.
Export orientation: While domestic demand remained small initially, Chinese manufacturers exported to Germany, Spain, Italy, and the U.S., capturing growing markets.
8.4.2 The Casualties
The rise of Chinese manufacturing devastated competitors:
- Germany: Q-Cells, once the world’s largest cell manufacturer, filed for bankruptcy in 2012
- United States: Solyndra (thin-film) failed in 2011, becoming a political controversy; SunPower, First Solar survived but struggled
- Japan: Sharp, Kyocera, and Sanyo lost market share; Sanyo was acquired by Panasonic
By 2012, module prices had fallen 80% from 2008 levels. Chinese manufacturers, operating on thin margins but massive scale, had transformed the industry’s economics.
8.4.3 Trade Conflicts
Western governments responded with tariffs:
United States: Anti-dumping and countervailing duties on Chinese panels (2012, 2014), eventually exceeding 100% on some products. Chinese manufacturers responded by shifting assembly to Southeast Asia.
European Union: Tariffs and minimum prices negotiated with Chinese exporters (2013-2018), eventually abandoned.
India: Tariffs on Chinese cells and modules (2017-present), attempting to build domestic manufacturing.
These trade measures had limited effect. Chinese dominance persisted, and global prices continued falling. The supply chain had become so integrated—with polysilicon, wafers, and cells crossing borders before final assembly—that tariffs often raised costs without reviving domestic manufacturing.
8.5 The Current Era: Grid Parity and Beyond (2015-Present)
8.5.1 Subsidy-Free Solar
By 2015, utility-scale solar had achieved “grid parity” in sunny regions—it could compete with wholesale electricity prices without subsidies. By 2020, solar was the cheapest source of new electricity generation in most of the world.
The policy landscape shifted accordingly:
Auctions: Many countries replaced FITs with competitive auctions, letting developers bid for the lowest price at which they’d build solar projects. Auction prices fell dramatically:
- UAE (2016): $0.03/kWh
- Chile (2017): $0.02/kWh
- Portugal (2020): $0.013/kWh
- Saudi Arabia (2021): $0.010/kWh
Corporate procurement: Large corporations (Google, Amazon, Apple) signed power purchase agreements directly with solar developers, bypassing utility procurement.
Self-consumption: In countries with high retail electricity prices (Germany, Australia), solar became attractive even without selling to the grid—just offsetting purchases was valuable enough.
8.5.2 China’s Domestic Boom
Having built global manufacturing dominance, China turned its solar industry inward. The 13th Five-Year Plan (2016-2020) targeted 105 GW of solar capacity; the actual result exceeded 250 GW. By 2024, China had more installed solar capacity than the rest of the world combined.
The scale is staggering. China installed 216 GW of solar in 2023 alone—roughly equal to the entire world’s cumulative installation through 2015. China now installs in one month what Germany installed in its entire decade of FIT-driven growth.
China solar installations (2023): 216 GW At 6 W/m2 effective power density: 36,000 km2 of solar farms That’s roughly the area of Taiwan, or about 0.4% of China’s land area.
At 1,500 kWh/kW/year capacity factor: \[216 \text{ GW} \times 1,500 \text{ h} = 324 \text{ TWh/year}\]
China’s electricity consumption: ~8,500 TWh/year Solar’s contribution: ~4% of electricity (growing rapidly)
China alone is now installing more solar capacity per year than global models projected for 2030 just a decade ago.
8.6 Technology Is Not Product
One of the deepest lessons from solar history is the distinction between a technology that works in a laboratory and a product you can buy at scale. The gap between them is typically 15-20 years:
| Stage | Duration | Key Risk |
|---|---|---|
| Lab → Pilot | 2-3 years | Yield collapse |
| Pilot → Factory | 3-5 years | Cost overruns |
| Factory → Market | 2-3 years | Customer trust |
| Market → Dominance | 10+ years | Incumbent catch-up |
Crystalline silicon had a 40-year head start on this pipeline (from Bell Labs in 1954 to commercial relevance in the 1990s). CIGS and other thin-film technologies tried to leapfrog, but discovered that you cannot shortcut the product pipeline. A cell that works brilliantly in a cleanroom fails at scale when yield drops from 95% to 90%, and that 5% difference can destroy the business case.
This distinction also explains why “boring” incremental improvements (thinner wafers, diamond wire saws, PERC rear passivation) outperformed “revolutionary” new materials (CIGS, amorphous silicon). Incremental improvements ride the existing production pipeline; revolutionary approaches must build a new one from scratch.
8.7 Five Limits That Weren’t
In 2008, the consensus view held that crystalline silicon faced several fundamental barriers. Every one of them turned out to be an engineering problem, not a physics problem:
1. Polysilicon supply: The Siemens process was slow, energy-intensive, and capacity-limited. New plants took 4 years to build. But China built 50+ fluidized bed reactor (FBR) plants, crashing prices from $400/kg to under $10/kg.
2. Indirect bandgap: Silicon absorbs light weakly compared to GaAs or CdTe, requiring thick wafers. But light-trapping techniques (texturing, PERC rear reflectors) enabled wafer thinning from 300 μm to 150 μm (and now below 120 μm) without efficiency loss.
3. Kerf loss: Sawing wafers from ingots wasted ~50% of the silicon as dust. Diamond wire saws reduced kerf from 180 μm to 50 μm, cutting waste from 50% to 25%.
4. Slow crystal growth: Czochralski pulling was a batch process. Manufacturers scaled to larger ingots (6” → 8” → 12”), longer pulls (1 m → 6 m), and automated the process, reducing labor by 80%.
5. Energy payback: Critics argued PV took 2-4 years to repay its embodied energy. But materials per watt fell 75%, electricity per kg of silicon fell 50%, and efficiency rose 60%. Energy payback dropped to 0.5-1 year.
The crucial insight: “Fundamental limits are sometimes just engineering problems waiting for scale.” The learning curve didn’t just reduce costs; it systematically demolished barriers that experts had called permanent.
8.8 The Subsidy Landscape
The solar industry’s growth occurred within a broader energy subsidy context that is often misunderstood:
G20 fossil fuel subsidies (2023): $535 billion. Renewable subsidies: $168 billion. The ratio is 3:1 in favor of incumbents.
IMF broader measure (including health costs and climate externalities): $7 trillion/year global fossil fuel subsidies, or 7.1% of global GDP.
US historical subsidies (1950-2016): Fossil fuels received $666 billion, nuclear $85 billion, and renewables $50 billion. Solar’s rapid cost decline occurred despite receiving far less cumulative support than its competitors.
Who paid for the learning curve? Germany spent €100B+ on feed-in tariffs from 2000-2020. China invested $47B+ in state support from 2010-2018 (plus subsidized land, electricity, and loans). The US IRA projects $421B from 2025-2034. Japan spent ~$90B from 2012-2020. The irony: Germany paid to create the market, and Chinese manufacturers captured the production.
8.9 Lessons from the History
8.9.1 What the Learning Curve Required
The 99% cost reduction from 1976 to today was not automatic. It required:
Early protected markets: Space applications (1960s-70s), Japanese rooftops (1990s-2000s), German FIT (2000s-2010s)
Sustained demand: Stop-and-go policy kills learning curves. Germany’s consistent growth enabled manufacturing investment.
Manufacturing scale: Cost reductions required not just technological improvement but factory-level learning and scale economies.
Competition: Multiple manufacturers competing on cost and efficiency accelerated improvement.
Global supply chains: Specialization (polysilicon in China, equipment in Europe, demand wherever) optimized the system.
8.9.2 Policy Design Principles
Comparing successful and unsuccessful solar policies reveals patterns:
Degression works: Declining subsidies create urgency, reduce windfall profits, and align incentives with cost reductions.
Duration matters: Long-term commitments (Germany’s 20-year guarantees) provide investment certainty; short-term programs create boom-bust cycles.
Simplicity aids adoption: Germany’s FIT was simple—generate power, get paid. Complex schemes with multiple conditions reduce uptake.
Equity requires attention: Policies perceived as unfair to non-participants face political backlash.
8.9.3 The Smil Caution
Vaclav Smil has repeatedly noted that even this extraordinary success—99% cost reduction, exponential deployment—has not fundamentally transformed the global energy system. Solar provides about 5% of global electricity and less than 2% of total primary energy.
His point: the history of solar demonstrates both what is possible (remarkable technological progress under the right conditions) and what remains (the sheer scale of the fossil fuel system to be displaced). Optimism about technology should not become optimism about timelines.
Solar’s history is a case study in Policy → Outcome, but also in the feedback loops that connect stages:
- Japan’s subsidies created demand that drove Technology improvement
- Germany’s FIT created a massive market that enabled Products (commercial-scale manufacturing)
- China’s industrial policy optimized the Technology → Product pipeline
- Competitive auctions revealed that Outcomes (cheap solar) could emerge without ongoing subsidies
The Framework operates in both directions. Early policies created outcomes that changed what subsequent policies were possible. By 2020, supporting solar no longer required justification—it was simply the cheapest option.
8.10 The Unfinished Story
8.10.1 Supply Chain Concentration
China’s dominance creates vulnerabilities:
- Polysilicon: 80% from China, with ~45% from Xinjiang (labor practice concerns)
- Wafers: 97% from China
- Cells: 85% from China
Western countries are attempting to rebuild domestic manufacturing:
- U.S. Inflation Reduction Act (2022): massive production tax credits for domestic manufacturing
- European Solar Act: targets 40% domestic manufacturing by 2030
- India PLI scheme: production-linked incentives for domestic cells and modules
Whether these efforts will succeed—or whether China’s advantages are too entrenched—remains uncertain.
8.10.2 Integration Challenges
As solar penetration increases, new challenges emerge:
- Grid stability: Variable output requires backup or storage
- Curtailment: Excess midday solar may go unused
- Value deflation: Adding more solar reduces the value of each additional kWh
These are Module 4 topics, but they’re integral to solar’s future. The next chapter examines current policy frameworks and the transition to wind energy.
8.11 Key Concepts
- Niche markets: Early development in protected markets (space, then rooftops)
- Feed-in tariffs: Guaranteed prices that created demand and enabled scale
- Degression: Declining subsidies that aligned incentives with cost reductions
- Industrial policy: Chinese state support that captured manufacturing dominance
- Grid parity: When solar competes without subsidies
- Supply chain concentration: Risks of depending on single-country production
8.12 Exercises
Subsidy calculation: Germany installed 7.6 GW of solar in 2012 at an average FIT of €0.25/kWh. If these systems operate for 20 years at 1,000 kWh/kW/year, what is the total subsidy commitment? (Assume wholesale electricity value of €0.04/kWh.)
Learning curve verification: Use the data points (1976: $106/W at 0.001 GW; 2024: $0.11/W at 1,800 GW) to calculate the learning rate. Does it match the claimed 24%?
Policy comparison: Japan’s program ran from 1994-2005, installing 1.4 GW at declining subsidies. Germany’s FIT from 2000-2012 installed 32 GW at higher subsidies. Estimate the subsidy per kWh of lifetime generation for each program (assume 1,000 kWh/kW/year for Japan, 900 kWh/kW/year for Germany).
Trade-off analysis: A country can support solar through: (a) FITs at €0.15/kWh for 20 years, or (b) upfront capital grants of €500/kW. Compare the total subsidy cost, risk allocation, and administrative complexity.
Supply chain security: If the U.S. wants to source 50% of its solar panels domestically by 2030 and deploys 50 GW/year, how much domestic manufacturing capacity is needed? What are the barriers to achieving this?
Historical counterfactual: Imagine Germany had invested €100 billion in nuclear rather than solar FITs (2000-2015). At €5,000/kW for nuclear (optimistic for Europe), how much nuclear capacity could have been built? At 90% capacity factor, how much electricity would it generate annually? How does this compare to the solar capacity built? What are the tradeoffs?
This chapter exemplifies Reduction to Understanding: starting from observed Outcomes (the solar cost revolution) and tracing backward to the Policies, Products, Technologies, and Principles that produced them.
The analysis reveals that technological Principle (the photovoltaic effect) was necessary but not sufficient. Technology development (efficiency improvements), Product commercialization (manufacturing scale), and Policy support (FITs, industrial policy) were all required. Remove any link and the chain breaks.
For the Energy RPG in Module 5, this history offers lessons: policy design matters enormously, equity considerations can’t be ignored, and even successful policies take decades to transform energy systems.