Claims that US solar is slowing usually rest on a single metric read out of context. This guide shows how to check capacity, generation and demand figures yourself, and how to tell a genuine slowdown from a statistical artefact.
Gather these before you start
You need three things: a data source, a set of definitions, and a spreadsheet.
For data, the main public repository is the US Energy Information Administration, which publishes monthly electricity statistics and a short-term outlook covering generation, capacity and consumption. The Federal Energy Regulatory Commission publishes periodic summaries of newly installed generating capacity. Grid operators and national laboratories publish information on projects waiting to connect to the grid. Industry associations publish their own installation totals, which use different boundaries and are worth treating as a separate series rather than a cross-check.
For definitions, write down what each column in your chosen dataset actually measures before you use it. Capacity, generation, additions, net additions, utility-scale and small-scale are all distinct concepts, and most disagreements about solar’s trajectory come from mixing them.
For tools, any spreadsheet will do. You are doing arithmetic on published tables, not modelling.
Separate capacity additions from electricity generated
Capacity is the maximum output a fleet of panels could produce under reference conditions, measured in megawatts. Generation is the electricity actually produced over a period, measured in megawatt-hours. They move together, but not in step.
A solar farm connected in November contributes almost nothing to that year’s generation total and almost all of its capacity. Capacity added late in one year shows up as generation growth in the next. This lag alone can make capacity and generation series appear to tell opposite stories about the same fleet.
Decide which question you are asking. If you want to know how much fossil generation solar is displacing, use generation. If you want to know how fast the industry is building, use capacity. Do not use one as evidence about the other.
Check whether rooftop solar is counted
Most electricity statistics distinguish utility-scale solar, meaning plants above a size threshold that sell into the grid, from small-scale or distributed solar, meaning rooftop and other behind-the-meter systems.
The two are often collected by different methods. Utility-scale output is metered and reported directly. Small-scale output is usually estimated, because the electricity is consumed on site and never crosses a meter the grid operator reads. Some published tables include only the utility-scale series, some combine both, and the combined series may be revised more heavily.
Before comparing any two figures, confirm they use the same boundary. A headline about slowing solar built on utility-scale data alone describes a different population from one built on the combined series.
Compare absolute additions with percentage growth
A percentage growth rate is a ratio, and its denominator grows every year. A fleet that adds the same number of gigawatts annually will show a falling growth rate indefinitely, even though nothing about the build rate has changed. A fleet that adds more gigawatts than last year can still show a lower growth rate.
So calculate both. Put annual capacity additions in one column and year-on-year percentage change in another. If additions are flat or rising while the percentage falls, you are looking at base effects, not a slowdown. If additions themselves decline over several consecutive periods, that is a real change in build rate and worth investigating.
State clearly which of the two any claim you read is based on. Many do not say.
Measure solar against the change in demand
Total electricity demand is the context that makes capacity numbers meaningful. The relevant comparison is not solar against the whole grid, but new solar generation against the year’s change in demand.
Ars Technica reports that renewables comfortably cover the changes in US electricity demand, with demand rising by 2 per cent in 2026. That framing is the one to reproduce with your own arithmetic: take the expected increase in consumption for the period, take the expected increase in generation from renewables, and compare them directly.
If new renewable generation exceeds demand growth, the fossil-fuelled share of the mix falls even if solar’s own growth rate is easing. If it falls short, something else covers the gap. This single comparison resolves more arguments than any capacity figure on its own.
Strip out seasonality and weather
Solar output follows the sun. Monthly generation peaks in summer and falls in winter in every year, regardless of how much capacity exists. Comparing consecutive months therefore measures the season, not the trend.
Compare the same month across years, or use rolling twelve-month totals. Both approaches remove the annual cycle without hiding genuine turning points.
Weather is the second layer. Cloud cover, dust and snow vary year to year and move output by amounts that can be mistaken for trend changes in a single month. A pattern that only appears in one month of one year is usually weather. Require several consecutive periods before treating a change as structural.
Look at the project pipeline before calling a turning point
Generation and capacity data are backward-looking. A plant producing electricity today was financed years ago. To say anything about the future, you need forward indicators: projects under construction, projects with signed connection agreements, and projects waiting in interconnection queues.
Queue data needs care, because a large share of queued projects are never built and some developers file multiple applications for the same site. Treat queue volume as an upper bound on future building, not a forecast. Projects already under construction are the more reliable indicator.
The size of any gap between the pipeline and recent installations is not something you can settle from the historical tables alone.
Avoid the mistakes that make trends look sharper than they are
The most common error is comparing series with different boundaries: utility-scale in one year, combined in another, or capacity measured in direct-current watts against capacity measured in alternating-current watts, a difference that shifts totals systematically.
The second is treating preliminary data as final. Recent months in most electricity datasets are estimates and get revised, often enough to change the direction of a short-run trend.
The third is using a single month, or a single quarter, as a trend. The fourth is ignoring the difference between announced projects, projects under construction and projects generating electricity; these are three different populations and press coverage moves between them freely.
The fifth is assuming that capacity growth and emissions reductions are the same measurement. They are related through the generation mix, not directly.
Skip this approach when the question is not about national electricity supply
This method answers questions about aggregate US electricity: how much is generated, from what, and how that changes. It is the wrong tool for several adjacent questions.
It will not tell you about the economics of an individual installation, where local prices, tariffs, incentives and roof conditions dominate and national data is irrelevant. It will not tell you about manufacturing, since panels installed in the US may be made anywhere and installation data says nothing about factory output.
It is also weak on grid-level questions. Whether a region can absorb more solar depends on transmission, storage and the shape of local demand, none of which appear in national generation totals. For those questions, use regional grid operator data instead.
Frequently asked questions
What is the difference between solar capacity and solar generation?
Capacity is potential output, measured in megawatts: how much a fleet of panels could produce at once under reference conditions. Generation is realised output over time, measured in megawatt-hours: how much electricity was actually produced. Capacity is a stock measured at a point in time, generation a flow measured over a period. A plant connected in December adds capacity to that year and generation to the next.
Where can I find official US solar electricity data?
The US Energy Information Administration is the primary public source, publishing monthly electricity statistics and periodic outlooks that cover generation by fuel, installed capacity and consumption. The Federal Energy Regulatory Commission publishes summaries of newly installed generating capacity. Both are free. Industry associations publish their own installation figures, which use different definitions and should be treated as a separate series rather than a verification of the federal numbers.
Does a falling growth rate mean solar is shrinking?
No. A percentage growth rate divides new additions by an existing base that gets larger every year. A fleet adding a constant number of gigawatts annually will show a steadily declining growth rate while building at exactly the same pace. Check absolute annual additions alongside the percentage. Only a sustained fall in absolute additions across several periods indicates that building has genuinely slowed.
How does solar growth relate to overall electricity demand?
Demand growth sets the bar that new generation must clear before it can displace existing sources. Ars Technica reports that renewables easily cover the changes in US demand, with consumption rising by 2 per cent in 2026. When new renewable generation exceeds the annual increase in demand, the share supplied by other sources falls. When it does not, the difference is met elsewhere in the mix.
Sources and further reading
- The US Energy Information Administration: monthly electricity statistics and short-term outlooks covering generation, capacity and consumption, with published methodology for the small-scale solar estimates.
- The Federal Energy Regulatory Commission: periodic reports on newly installed generating capacity by fuel type.
- Ars Technica: reporting on the pace of US solar growth and on renewables covering changes in electricity demand.
- Regional grid operators and national laboratories: interconnection queue and project pipeline data, useful as forward indicators with the caveats noted above.
Surfaced from the rss:arstechnica signal “us solar growth data”. AI-assisted draft, editorially reviewed.
