| Episodes | 70 |
| Mentions | 129 |
| Cited here | 42 |
| First — last | #44 — #705 |
| Top guests | Ari Gerstman, Martin Lorton, Paul Zawada |
| Related | power grid · solar panel · battery |
Solar power is the conversion of sunlight into usable energy, most commonly electricity generated by photovoltaic panels. Its deployment is governed primarily by economics: installed cost per watt, retail electricity prices, and tariff design determine whether systems pay back their cost, after which the energy is effectively free.[415][540] Because generation peaks hours before demand, integrating large amounts of solar into the grid creates operational problems around voltage compliance, forecasting, and storage that shape utility policy.[580][630]
Manufacturing
Crystalline silicon for solar cells is produced through an energy-intensive multi-stage chemical process. Sand is reduced with coal into a slag, which is dissolved in acid and reacted to produce silane gas; the gas is then run into reactors that deposit pure silicon onto long rods, and the deposited silicon is remelted and drawn into a boule.[52]
Cell characterization and power electronics
A solar cell is properly characterized by loading it and sweeping its output curve to find the peak power point, rather than by taking an open-circuit voltage reading.[565] Because a panel’s open-circuit voltage can exceed the nominal voltage of the storage cell it charges, a buck converter is used to manage the difference, for example trickle-charging a six-volt gel cell from a panel whose output reaches eight or nine volts.[561]
System design and siting
Roof geometry and shading
Before quoting a residential installation, a roof-modelling tool is used to map usable area, energy per square area, and tree shading from a 3D model; complex multi-facet roofs can be ruled out on geometry alone because panels are rectangular.[555] Roof orientation determines which part of the day an array produces: a west-facing main roof yields only afternoon sun, so a smaller but south-facing, unshaded garage roof can be the better mounting surface.[688] Tree shading has historically been the blocking constraint for residential installs in wooded neighborhoods, but newer panel-level electronics make shaded roofs workable.[206]
Where a local utility will not buy exported energy and interconnection paperwork is heavy, the rational design choice is a system configured never to feed the grid at all.[206]
Self-consumption and load scheduling
When the export tariff is negligible, appliances with timers should be scheduled into daylight hours so that generation is self-consumed rather than exported; unused daytime output is effectively wasted.[368] Deferrable loads such as home servers can be gated to run only on surplus generation, either on a simple daytime timer or on an insolation threshold that switches the load in once output passes a set level.[296] Timer-scheduled loads fail on overcast days, however, because the timer has no knowledge of actual generation and the load ends up draining the house battery instead of consuming surplus.[683] For the same reason, when a system already generates a large daytime surplus, doubling the energy a daytime appliance consumes barely moves the payback period, since the extra consumption only displaces export sold at a very low rate.[610]
A solar-diverting electric-vehicle charger senses export with current clamps and starts charging the car once surplus exceeds a set threshold, regulating the car’s draw by varying the pilot PWM duty cycle; the resulting control loop has a delay of a few seconds before the vehicle compensates.[612][654]
Storage
For cold-weather solar deployments, a six-volt gel cell trickle-charged from the panel is a more robust storage choice than lithium-ion.[561] Lithium-ion buffering is used in milder conditions: a solar-powered sensor sealed inside a septic tank, with its panel on a cable to the exterior, carries roughly two to three weeks of reserve at hourly 3G uploads with no sun.[355]
Home battery scale is small relative to vehicle packs: a 12.5 kilowatt-hour home battery is roughly equivalent to a single Powerwall, and a 25 kilowatt-hour bank remains small against an electric car pack of 38 to 58 kilowatt-hours.[696] Even so, a relatively small home battery is enough to bridge the evening dip in a solar-equipped home, with one system stopping grid imports around noon and not drawing from the grid again until 8 or 9 p.m.[702] A home battery only pays without solar if the meter supports time-of-day tariffs; on a flat-rate meter, day and night power cost the same and there is no arbitrage to capture.[580] A large array with battery storage can still leave twenty kilowatt-hours a day of unusable surplus coming out of winter, showing that storage sized for the house does not absorb all generation.[677]
At grid scale, a solar-powered factory that must run continuously forces its owner into grid-scale storage, since production and round-the-clock demand do not coincide.[210] Solar oversupply is fundamentally a storage shortage: with enough grid storage, such as pumped hydro, the same generation would not present a problem.[702]
Economics and tariffs
Feed-in tariffs come in net and gross variants: net pays only for surplus exported after the house consumes what it needs, while gross pays for all generation regardless of on-site use.[150] Tariff levels directly steer demand. A 60-cent-per-kilowatt-hour feed-in tariff plus government subsidies drove Australian rooftop solar to about 35 percent of homes, and Australian uptake later reached roughly 40 percent of houses, the highest in the world, while electric vehicles remained about 0.1 percent of car sales there.[606][524] Where grid electricity costs only eight or nine cents per kilowatt-hour, rooftop solar has no economic driver; high retail prices are what pushed Australian rooftop penetration past 13 percent of homes.[249] Australian feed-in tariffs later inverted from paying about three times the retail rate to paying about one quarter of it, flipping the economics from exporting to self-consuming daytime generation.[127]
Stacked incentives can dominate residential economics. One Oregon system had about a third of its cost covered by an Energy Trust incentive and slightly more than another third by state and federal credits, leaving 4,300 out of pocket; the system was guaranteed 4.8 megawatt-hours per year and delivered 5.26 megawatt-hours in its first full year.<sup><a href="#ref-112" title="Ep 112: An Interview with Bob Simpson - Ardent Automotive Artisan (September 9, 2012)">[112]</a></sup> An Australian 1.5-kilowatt system cost about four thousand dollars after a large early subsidy was wound back, while a top-of-the-line three-kilowatt system bought around 2015 cost 5,000 after state and federal rebates and paid for itself in about five years.[127][606] Payback period scales with equipment tier: a low-cost residential system pays back very quickly, while a top-quality system takes longer, though still on the order of five years.[540] In a US state with few incentives, quoted residential solar came back with an eighteen-year payback, which is the practical reason such installs get abandoned.[606]
Falling headline solar spending worldwide can mask growing deployment, because the cost per installed watt dropped; the per-watt figure, not total dollars, is the number to track.[343] Once solar falls below a threshold cost per watt, deployment stops depending on policy and becomes a pure market outcome.[329]
Grid integration
Conventional generators earn disproportionately at peak, when extra turbines are brought online on hot days, so distributed solar that shaves those peaks undermines the revenue model on which the grid is financed.[150] Wholesale price and demand traces show sharp evening spikes around 5 to 5:30 p.m., as households return and switch on cooktops just as solar output collapses; that is when expensive peaking plant must be started.[692] The resulting supply profile is known as the duck curve: solar supply peaks hours before demand peaks, and managing the mismatch is a principal justification for deploying storage.[630]
Very high rooftop penetration creates an operational problem: midday export pushes distribution voltages toward or past the compliant limit, which is why utilities seek the ability to curtail residential systems.[580] When a feeder carries more distributed solar than local load can absorb, curtailment of that generation is the first tool available to a utility.[630] Getting permission to install a residential array can be easy while authorization to operate is the hard part, because the utility’s underlying requirement is retaining the ability to shut the system off when a feeder has more supply than load.[630] One plausible utility arrangement pays homeowners a monthly premium plus a negotiated rate in exchange for contractual rights over their battery, such as requiring at least 50 percent state of charge at 6 p.m. in summer and unrestricted draw from it.[630]
Exported rooftop power travels locally: it reaches a neighbor with a short hop because both premises sit on the same distribution feeder, whereas moving power between blocks requires routing up to the substation and back down.[630] For a household without panels but with solar-equipped neighbors, charging an EV at midday draws from the most local, lowest-impedance source rather than from distant generation.[666]
High penetration also changes planning. Aggregate household demand is highly predictable, but solar forces utilities to fold weather forecasts into generation planning, adding a source of error that did not previously exist.[610] Solar is an inverter-based resource, as is modern wind: large wind turbines do not spin synchronously with the grid and instead generate into DC and feed the grid through an inverter.[583] Solar-dedicated policy has followed the oversupply problem: Australia’s surplus midday solar prompted a two-billion-dollar national subsidy scheme to put a battery in every home, on the reasoning that the grid has nowhere else to absorb the excess.[702] One long-time renewables advocate, having previously worked through reports concluding a fully renewable Australian grid was achievable, reversed position and came to regard nuclear as necessary for baseload, noting a ten-year ramp and no domestic expertise.[610]
Applications
Energy harvesting and remote sensing
Among ambient energy sources, a few square inches of solar cell with battery backup is the most successful harvesting approach, with thermoelectric generation second and mechanical harvesting rarely working out; the right choice is highly application specific, so a hot pole-mounted transformer argues for thermoelectrics.[232] Battery-plus-solar sensors embedded in roadway concrete avoid cutting asphalt for wiring and can run five years before needing maintenance, which makes fire-and-forget deployment practical.[179]
Spacecraft
On a small satellite, the available solar area is the hard outer bound on the power budget, so the design method is to drive every subsystem as low-power as possible and then trade away duty cycle, for instance by not sampling over certain regions.[518] For commercial spacecraft, solar is effectively the only practical power source, because flying a radioisotope thermoelectric generator requires presidential-level authorization and the plutonium supply is constrained by nuclear arms treaties.[701]
Vehicles and wearables
Roof-mounted panels on a car top out around 800 watts, which yields too little daily energy to sustain continuous driving even with full sun in the Australian outback.[627] The economics of solar-on-vehicle were illustrated by Aptera’s bankruptcy: an otherwise ordinary EV with roof panels priced around a quarter of a million dollars had no viable market.[619] At the smallest scale, wearable-sized solar cells cannot collect enough energy to meaningfully recharge a phone battery, making solar-powered wristband power banks an energy-budget impossibility rather than an engineering challenge.[233]
Water heating
Solar thermal panels that circulate water through roof plumbing add cost and complexity and shift the failure mode from an electrical callout to a plumbing one.[674] A heat-pump hot water system uses so little energy—about three kilowatt-hours a day, roughly a tenth of what a domestic PV array makes even on a winter day at thirty kilowatt-hours—that the case for solar thermal is removed.[674]
Deployment milestones
The 1970s push for solar failed principally on cost; the durable argument for solar is the economic one, that the system earns its cost back and then supplies power for free.[415] Germany became the first country to briefly supply half its electricity from solar, a milestone reached despite unremarkable solar irradiation.[205] Even with costs dropping quickly, solar was still under five percent of world energy supply as of the mid-2010s, a scale check against enthusiastic growth figures.[439]