The Solar Fact Check
Solar electricity became the cheapest energy source in history. It now generates 5.5% of U.S. power. The remaining 94.5% is the story the cost charts don't tell — and the four documented facts that explain why the gap between falling prices and grid dominance is larger than the headlines suggest.
The cost of solar electricity has undergone the most dramatic price collapse in the history of energy technology. IRENA's 2023 renewable power generation costs report documents a 90% decline in utility-scale solar photovoltaic LCOE between 2010 and 2023 — from roughly $0.38 per kilowatt-hour to $0.049 per kilowatt-hour globally. Lazard's 2024 LCOE analysis puts U.S. utility-scale solar at $24 to $38 per megawatt-hour before incentives. That is cheaper than the operating cost of existing coal plants in most U.S. regions, and below the fuel cost of new gas generation.
This is not a projection. It is documented data. The price decline happened, it is ongoing, and it is real. Any analysis of solar energy that does not begin with this fact is not analysis — it is advocacy in the other direction.
The 5.5% figure is also real. In 2024, solar generated approximately 5.5% of total U.S. electricity, according to EIA monthly generation data. This represents a 55-fold increase from 2010, when solar contributed less than 0.1%. The trajectory is genuine. So is the gap between the trajectory and the destination.
What Capacity Factor Actually Means
Electricity grids are not designed to answer the question "what is the cheapest source over a year." They are designed to answer the question "what source can reliably generate power at the moment it is needed." Those are different questions, and capacity factor is the metric that connects them.
Capacity factor is the ratio of actual electricity generated to the maximum theoretical output at full rated power, measured over a year. A 100 MW natural gas plant running at a 55% capacity factor produces about 55 MW of equivalent continuous power. Nuclear plants run at roughly 93% capacity factor — they produce nearly their full rated output around the clock, day and night, 335 days a year. Solar PV in the continental United States averages 15% to 25% capacity factor depending on location, because panels generate power only when the sun shines, and output varies with angle, cloud cover, season, and time of day.
This does not make solar "bad." It makes solar a different kind of resource than dispatchable generation. A 1-gigawatt solar installation produces, over a year, the same amount of electricity as roughly a 180–250 MW gas plant running continuously. The cost per megawatt-hour can be lower while the cost per megawatt of reliable capacity remains much higher, because reliability requires either storage or backup generation to cover the hours when the sun does not shine.
The Duck Curve: A Grid Operator's Documented Problem
In 2013, the California Independent System Operator published a study projecting what high solar penetration would do to California's net load curve — the demand grid operators must meet from non-solar sources. The resulting graph, which resembles a duck in cross-section, became the defining image of solar integration challenges.
The duck curve problem has two parts. First, as solar output peaks midday, it suppresses the apparent demand from the grid. Non-solar generators — gas plants, nuclear, hydro — must reduce output or risk frequency instability, which causes curtailment: solar energy that is generated but cannot be used. California curtailed approximately 3% of its potential solar generation in 2023, with curtailment concentrated on spring days when solar output is high and heating and cooling demand is low.
Second, and more operationally demanding, is the evening ramp. When the sun sets, solar output falls rapidly while residential demand spikes as people return home. CAISO has documented evening ramps requiring over 15,000 megawatts of new generation to come online within a three-hour window. Natural gas plants, which can ramp at 50–100 MW per minute, provide most of this balancing capacity. The irony is measurable: high solar penetration increases the operational value of natural gas peaker plants, because the grid needs something that can go from zero to full output in minutes when the sun goes down.
"The deeper irony of high solar penetration is that it has increased, not decreased, the operational indispensability of fast-ramping natural gas. CAISO's evening ramp requirement grew from roughly 7,000 MW in 2012 to over 15,000 MW in 2023 — a direct consequence of the solar capacity installed during the same period."
The China Supply Chain Problem
The International Energy Agency's 2024 solar supply chain report documents that China currently produces approximately 80% of global solar panels. More specifically: China accounts for 97% of global polysilicon production, 97% of silicon wafer manufacturing, 85% of solar cell production, and 80% of completed module manufacturing. Every other node in the supply chain is a rounding error.
This concentration was not accidental. A Lawrence Berkeley National Laboratory analysis found that Chinese solar manufacturers received approximately $130 billion in government subsidies between 2011 and 2022 — including preferential land, subsidized electricity, state-backed financing, and direct production incentives. The result was a cost advantage that eliminated most non-Chinese solar manufacturers. German solar manufacturer Q.Cells, once the world's largest, entered insolvency in 2012. U.S. manufacturer Solyndra became a political scandal in the same year for the same structural reason: Chinese manufacturers were producing panels at costs that undercut any Western producer operating without equivalent government support.
The Uyghur Forced Labor Prevention Act, enacted in 2022, created a rebuttable presumption that goods produced in Xinjiang — where most of China's polysilicon is manufactured — involve forced labor and are therefore banned from U.S. import. The practical effect has been to create a documentation requirement, not a supply chain shift. Most analysts estimate that 40–50% of solar panels installed in the United States in 2024 still contain components originating from Xinjiang, either directly or through third-country transshipment through Vietnam, Malaysia, and Thailand.
The $369 Billion Bet
The Inflation Reduction Act, signed August 2022, allocated approximately $369 billion in climate and clean energy provisions over ten years, according to the Congressional Budget Office's initial scoring. The centerpiece is an expanded Investment Tax Credit (ITC) of 30% for solar installations, with adders for domestic content (10%), energy communities (10%), and low-income communities (10%). The production tax credit structure extends through 2032 and is technology-neutral after 2025.
The IRA's domestic content provisions are the most directly relevant to the supply chain problem. To claim the domestic content adder, projects must use U.S.-manufactured steel and iron, and meet escalating thresholds for domestically produced components — starting at 40% in 2023 and rising to 55% by 2027. The intent is to rebuild U.S. solar manufacturing capacity. First Solar, which manufactures cadmium telluride thin-film panels in Ohio, has benefited directly: the company's stock price rose roughly 60% in the year following IRA passage, and it announced expansion to a fourth U.S. manufacturing facility. But for the crystalline silicon market — roughly 95% of installed solar — domestic supply chain reconstruction requires building out polysilicon, wafer, cell, and module manufacturing essentially from zero, against a Chinese industry that has spent 15 years optimizing every node.
The Lawrence Berkeley National Laboratory's "Utility-Scale Solar 2023" report documents 2,600 gigawatts of solar capacity in the U.S. interconnection queue — more than twice current total U.S. generating capacity. The queue number reflects both genuine project development and speculative applications. LBNL estimates that roughly 20–25% of projects in the queue will ultimately reach commercial operation. The grid infrastructure required to integrate what does get built — new transmission lines, storage facilities, and grid-forming inverter standards — is a separate constraint that the LCOE number does not capture.
What the Data Shows and What It Doesn't
The four documented facts about solar tell a coherent story that is more interesting than either the promotional version or the dismissive version. Solar is genuinely the cheapest electricity source by LCOE in most markets. It is genuinely intermittent in a way that requires either storage or backup. China genuinely controls the supply chain in a way that creates strategic risk. And $369 billion is a genuine government attempt to reshape both the economics and the supply chain simultaneously.
What the data does not resolve: whether long-duration storage will develop fast enough and cheaply enough to make high solar penetration economically viable without natural gas backup. The current generation of lithium-ion battery storage is optimized for 4-hour discharge — useful for the evening ramp but not for multi-day low-solar events. Iron-air batteries, pumped hydro, and other long-duration storage technologies are in various stages of development and early deployment. Form Energy has announced utility-scale iron-air installations. Pumped hydro remains geographically constrained. The math on a 100% solar-and-storage grid has not been demonstrated at scale in a high-latitude, high-load environment. Germany's Energiewende, the most ambitious attempt to date, has resulted in the highest retail electricity prices in the developed world and continued reliance on imported natural gas for baseload until Russian supply disruption forced an emergency coal extension in 2022.
The technology story is real. The grid integration story is more complicated. Both are documented. The investigation maps what the evidence shows.