6 Realtime Sun—Principles
“The Sun delivers more energy to Earth in one hour than humanity consumes in a year. The engineering challenge is not supply but conversion.”
Chapter 4 traced the “long path” from sunlight to energy: nuclear fusion produces radiation that travels to Earth, where photosynthesis captures ~0.5-2% of it, geological processes preserve ~0.1% of that biomass, and after millions of years we extract it as fossil fuel. The entire chain delivers power densities of 1,000-10,000 W/m2 at extraction, but consumes an irreplaceable inheritance.
This chapter examines the “short path”: capturing sunlight directly, in real time. This approach skips photosynthesis and geological concentration entirely, but it means working with the solar flux as it arrives, averaged and variable, without the concentration that time provides.
6.1 The Physics of Sunlight Revisited
6.1.1 The Solar Constant
As derived in Chapter 4, the solar constant is the power density of sunlight at Earth’s orbital distance:
\[S_0 = 1,361 \text{ W/m}^2\]
This value comes from the inverse square law applied to the Sun’s luminosity:
\[I = \frac{L_\odot}{4\pi d^2} = \frac{3.83 \times 10^{26} \text{ W}}{4\pi (1.496 \times 10^{11} \text{ m})^2} = 1,361 \text{ W/m}^2\]
The solar constant varies slightly (±0.1%) with the 11-year solar cycle and more substantially (±3.4%) with Earth’s elliptical orbit (closest in January, farthest in July).
Earth’s cross-sectional area (the circular shadow it would cast): \[A = \pi R_E^2 = \pi \times (6.37 \times 10^6)^2 = 1.27 \times 10^{14} \text{ m}^2\]
Total solar power intercepted: \[P_{total} = S_0 \times A = 1,361 \times 1.27 \times 10^{14} = 1.73 \times 10^{17} \text{ W}\]
That’s 173,000 TW, approximately 10,000 times current human power consumption (~18 TW).
But this power is spread across Earth’s surface area, which is four times the cross-section (full sphere vs. disk): \[\text{Average flux} = \frac{S_0}{4} = 340 \text{ W/m}^2\]
And roughly 30% is reflected back to space (Earth’s albedo), leaving: \[\text{Absorbed average} = 340 \times 0.7 = 238 \text{ W/m}^2\]
This is the starting point for both pathways: photosynthesis (Chapter 4) captures 0.5-2% of this, while photovoltaics can capture 15-25%.
6.1.2 The Solar Spectrum
Sunlight is not monochromatic but spans a range of wavelengths. The Sun’s surface temperature (~5,778 K) determines this spectrum through Planck’s blackbody radiation law:
\[B(\lambda, T) = \frac{2hc^2}{\lambda^5} \frac{1}{e^{hc/\lambda k_B T} - 1}\]
The result is a distribution peaking in the visible range (around 500 nm, green-yellow light) with substantial energy in both ultraviolet (shorter wavelengths) and infrared (longer wavelengths).
The spectral composition matters enormously for solar conversion technologies. The energy of a photon is: \[E = h\nu = \frac{hc}{\lambda}\]
where \(h\) is Planck’s constant, \(c\) is the speed of light, and \(\lambda\) is wavelength. This means:
- Ultraviolet photons (λ < 400 nm): E > 3.1 eV
- Visible photons (400-700 nm): E = 1.8-3.1 eV
- Near-infrared photons (700-1400 nm): E = 0.9-1.8 eV
The distribution of photon energies sets fundamental limits on photovoltaic conversion, as we’ll see shortly.
6.1.3 Atmospheric Effects
The solar spectrum reaching Earth’s surface differs from the extraterrestrial spectrum due to atmospheric absorption and scattering:
Rayleigh scattering: Air molecules scatter short wavelengths more than long ones (proportional to λ-4). This is why the sky is blue and why sunsets are red—at low sun angles, light passes through more atmosphere, scattering away more blue.
Ozone absorption: The ozone layer absorbs most ultraviolet radiation below 300 nm, protecting life from DNA-damaging radiation but also removing that energy from the available solar resource.
Water vapor and CO2 absorption: Infrared wavelengths show distinctive absorption bands where atmospheric gases absorb strongly. The solar spectrum at ground level has “notches” at these wavelengths.
Aerosols and particulates: Dust, smoke, and pollution scatter and absorb additional light, varying strongly with location and weather.
The cumulative effect is quantified by the Air Mass (AM) coefficient:
- AM0: Extraterrestrial (above atmosphere)
- AM1: Sun directly overhead (at zenith)
- AM1.5: Sun at 48° elevation, the standard reference for solar cell testing
- AM2: Sun at 30° elevation
At AM1.5, approximately 1,000 W/m2 reaches a surface perpendicular to the Sun on a clear day—the standard test condition for rating solar panels.
6.2 The Solar Resource
6.2.1 Geographic Variation
The solar resource varies enormously with location. Several factors combine:
Latitude: Higher latitudes receive sunlight at lower angles, spreading the same energy over larger areas. A surface at 60°N latitude receives only half the energy per square meter of one at the equator, even ignoring seasonal variation.
Season: Earth’s 23.4° axial tilt creates seasons. At high latitudes, winter days are short and the sun is low; summer days are long but never make up the annual deficit. Near the equator, seasonal variation is minimal.
Cloud cover: Humid and tropical regions often have persistent clouds that dramatically reduce surface irradiance. Counterintuitively, some equatorial rainforest regions receive less annual solar energy than mid-latitude deserts.
Altitude: Higher elevations have less atmosphere above them, reducing atmospheric losses. The Atacama Desert in Chile (high altitude, extreme aridity) has the world’s highest recorded solar irradiance.
6.2.2 Measuring the Resource
The solar resource is typically quantified as:
Global Horizontal Irradiance (GHI): Total solar radiation on a horizontal surface, including both direct beam and diffuse (scattered) components.
Direct Normal Irradiance (DNI): Solar radiation in the direct beam only, measured perpendicular to the sun. This is what concentrating solar systems can use—you cannot focus diffuse light.
Plane of Array (POA): Irradiance on a tilted surface, such as a solar panel. Tilting toward the equator captures more energy than a horizontal surface at high latitudes.
Excellent solar resources:
- Atacama Desert, Chile: GHI > 2,700 kWh/m2/year
- Sahara Desert: GHI > 2,400 kWh/m2/year
- U.S. Southwest: GHI > 2,200 kWh/m2/year
Moderate resources:
- Southern Europe: GHI = 1,400-1,800 kWh/m2/year
- U.S. Midwest: GHI = 1,500-1,800 kWh/m2/year
Poor resources:
- Northern Europe: GHI = 900-1,200 kWh/m2/year
- Pacific Northwest: GHI = 1,000-1,400 kWh/m2/year
In the U.S. Southwest (GHI ≈ 2,200 kWh/m2/year):
- Daily average: 2,200 / 365 = 6.0 kWh/m2/day
- In power terms: 6.0 kWh/day ÷ 24 h/day = 250 W/m2 average
In Germany (GHI ≈ 1,100 kWh/m2/year):
- Daily average: 3.0 kWh/m2/day
- Power terms: 125 W/m2 average
This 2:1 ratio might suggest solar is twice as good in Arizona as in Germany. But Germany installed more solar capacity than the U.S. for many years. Economics and policy matter as much as physics—a theme we’ll explore in Chapter 8.
6.2.3 Temporal Variation
Solar energy’s fundamental challenge is variability:
Daily cycle: Zero output at night, rising to peak near solar noon, falling again. On a clear summer day, the temporal pattern is highly predictable; clouds create rapid fluctuations.
Seasonal cycle: At 40°N latitude, winter daily insolation is about 40% of summer values. At 60°N, winter drops to 15% of summer. This creates fundamental matching problems between supply and demand.
Weather variability: Clouds can reduce output by 80-90% within minutes. Dust, snow cover, and atmospheric haze create additional unpredictability.
This variability is solar’s Achilles heel—or rather, it’s the reason why solar can’t simply replace fossil fuels kilowatt-for-kilowatt. We’ll address storage and grid integration in Module 4.
6.3 The Photovoltaic Effect
6.3.1 Semiconductors and Band Gaps
The photovoltaic effect—converting light directly to electricity—relies on the quantum mechanics of semiconductors. In a semiconductor like silicon, electrons exist in two distinct energy bands:
- Valence band: Lower energy states where electrons are bound to atoms
- Conduction band: Higher energy states where electrons can move freely
The band gap (\(E_g\)) is the energy difference between these bands. For silicon: \[E_g = 1.12 \text{ eV}\]
When a photon with energy greater than the band gap strikes the semiconductor, it can excite an electron from valence to conduction band, creating an electron-hole pair. This is the photovoltaic effect.
Critical insight: the photon must have sufficient energy. A photon with energy less than the band gap passes through (or is absorbed as heat) without creating electron-hole pairs. The semiconductor is transparent to sub-bandgap photons.
6.3.2 The Shockley-Queisser Limit
In 1961, William Shockley and Hans Queisser derived the fundamental efficiency limit for a single-junction solar cell. Their analysis identified three unavoidable loss mechanisms:
1. Sub-bandgap transmission: Photons with energy below \(E_g\) cannot be absorbed. For silicon (\(E_g = 1.12\) eV, corresponding to λ = 1,100 nm), about 19% of solar energy is in photons with wavelengths longer than 1,100 nm—this energy is lost.
2. Thermalization: Photons with energy above \(E_g\) create electron-hole pairs, but the excess energy (above \(E_g\)) is rapidly lost as heat. A 3 eV photon striking silicon creates one electron-hole pair at 1.12 eV; the remaining 1.88 eV becomes heat. Since the solar spectrum peaks around 2 eV, this loss is substantial—about 33% of incident energy.
3. Radiative recombination: At thermal equilibrium, some electron-hole pairs recombine and emit photons. This sets a theoretical limit even for perfect devices.
The Shockley-Queisser limit for a single junction under AM1.5 sunlight:
| Band gap (eV) | Maximum efficiency |
|---|---|
| 0.9 | 30% |
| 1.1 | 33% |
| 1.4 | 33% |
| 1.7 | 29% |
The optimal band gap is approximately 1.3-1.4 eV, where the tradeoff between transmission and thermalization losses is best balanced. Silicon at 1.12 eV is close but not quite optimal, with a theoretical limit of about 33%.
The Shockley-Queisser limit illustrates the Principle stage of our framework. Before any technology, product, or policy, physics establishes what is possible. A single-junction silicon cell cannot exceed ~29% efficiency under standard sunlight, regardless of engineering improvements.
Understanding this limit shapes Technology development. Since single junctions are limited, researchers pursued:
- Multi-junction cells (stacking semiconductors with different band gaps)
- Concentrating systems (reducing semiconductor area with cheap optics)
- Hot-carrier cells (harvesting thermalization energy)
- Multi-exciton generation (creating multiple electron-hole pairs per photon)
Each approach tries to circumvent different aspects of the Shockley-Queisser constraints.
6.3.3 Real-World Efficiency Losses
Even the Shockley-Queisser limit assumes a perfect device. Real cells have additional losses:
Reflection: Untreated silicon reflects ~35% of incident light. Anti-reflection coatings reduce this to <5%.
Recombination at defects: Crystal defects, grain boundaries, and surface states cause electron-hole pairs to recombine before contributing to current.
Resistance: Electrical resistance in the semiconductor and metal contacts causes I2R losses.
Shading from contacts: Metal grid lines that collect current block some incident light.
Temperature effects: Cell efficiency decreases ~0.4-0.5% per °C above 25°C. A cell at 60°C (common in field conditions) loses ~15% relative efficiency.
Commercial silicon cells achieve ~20-23% efficiency—about 65-70% of the Shockley-Queisser limit. Laboratory records exceed 26%, approaching the practical limit for single-junction silicon.
6.4 Comparing the Two Paths: PV vs. Photosynthesis
Chapter 4 showed that photosynthesis achieves only 0.5-2% efficiency due to:
- Only ~45% of wavelengths are usable (chlorophyll absorption spectrum)
- ~25% quantum efficiency losses
- ~30% carbon fixation efficiency (RuBisCO limitations)
- ~50% respiration losses
- Various practical factors
The Shockley-Queisser analysis shows why photovoltaics can do much better:
| Factor | Photosynthesis | Silicon PV |
|---|---|---|
| Usable wavelengths | ~45% | ~70% |
| Quantum efficiency | ~25% | ~90% |
| Additional losses | Respiration, carbon fixation | Thermalization, reflection |
| Overall efficiency | 0.5-2% | 20-25% |
Silicon PV is 10-40× more efficient than photosynthesis at converting sunlight to useful energy.
This efficiency advantage explains why the “short path” (direct solar conversion) can achieve higher power densities than the “long path” (biomass → fossil fuel), even without the geological concentration that gives fossil fuels their remarkable extraction rates.
If photosynthesis were 20% efficient instead of 1%, the economics of biofuels, biomass power, and even agriculture would be fundamentally different. The ~20× efficiency gap between plants and PV is not a matter of engineering; it’s built into the biochemistry of photosynthesis, optimized by evolution for survival and reproduction rather than energy conversion efficiency.
This is why direct solar (PV) makes more sense for electricity than the biological pathway, and why biofuels struggle to compete with electrification for most applications. The “long path” through photosynthesis wastes most of the incident sunlight before the energy even enters the biological system.
6.5 Beyond Single Junctions
6.5.1 Multi-Junction Cells
The most direct way to beat the Shockley-Queisser limit is to use multiple semiconductors with different band gaps, stacked on top of each other. Each junction absorbs a different part of the spectrum:
- Top cell (high band gap): Absorbs high-energy photons, transparent to low-energy
- Middle cell(s): Absorb intermediate energies
- Bottom cell (low band gap): Absorbs remaining low-energy photons
A triple-junction cell (2.0/1.4/1.0 eV) has a theoretical efficiency limit of ~49% under unconcentrated sunlight. Practical cells achieve 32-39% in the laboratory, primarily used in space applications where high efficiency justifies high cost.
6.5.2 Concentrating Photovoltaics (CPV)
Instead of covering large areas with semiconductors, concentrate sunlight using mirrors or lenses. This allows:
- Using expensive, high-efficiency multi-junction cells economically
- Increasing power per cell (reducing cell cost per watt)
- Concentration factors of 500-1000× are common
At 1000× concentration, only direct normal irradiance matters (you cannot focus diffuse light), limiting deployment to regions with high DNI. The systems require two-axis tracking to keep the sun focused on the cells.
CPV theoretical limits under concentration:
- Single junction at 1000×: ~37%
- Triple junction at 1000×: ~51%
Record efficiencies exceed 47% under concentration. But CPV remains a niche technology due to complexity and the remarkable cost reductions in conventional flat-plate PV, a story we’ll explore in Chapter 7.
6.5.3 Silicon Cell Specifications
A typical commercial silicon solar cell (6-inch, 156 mm × 156 mm) has the following electrical characteristics:
- Open-circuit voltage (Voc): ~0.70 V
- Short-circuit current (Isc): ~10 A
- Fill factor (FF): ~0.80
- Efficiency: 20-23%
- Power output: ~6 W per cell
A standard 72-cell module wires these cells in series, producing ~40 V and 400-600 W depending on cell efficiency. The fill factor measures how closely the cell’s actual power output approaches the theoretical maximum of Voc × Isc; real cells lose some power to series resistance and recombination losses.
6.5.4 Cell Technology Taxonomy
Different cell architectures trade off efficiency against manufacturing cost:
| Technology | Efficiency | Cost | Status |
|---|---|---|---|
| PERC (Al-BSF) | 22-23% | Low | Dominant (declining) |
| TOPCon | 24-25% | Medium | Rapidly scaling |
| HJT (Heterojunction) | 24-26% | Higher | Growing |
| IBC (Back Contact) | 25-26% | Highest | Niche (SunPower) |
| CdTe (thin-film) | 19-21% | Low | First Solar only |
| Perovskite | 26% (lab) | Unknown | Pre-commercial |
| Perovskite/Si tandem | 33%+ (lab) | Unknown | 2027-2030 expected |
The industry is transitioning from PERC to TOPCon, driven by a 1-2 percentage point efficiency gain at modest cost increase. The tandem cells (perovskite on silicon) represent the next frontier: LONGi achieved 34.85% efficiency in 2024, breaking single-junction limits. If stability and manufacturing challenges are resolved, tandems could reach commercial production by 2027-2030.
6.5.5 The Cost Revolution: Swanson’s Law
The price of solar cells has followed a remarkably consistent learning curve:
| Year | Module Price ($/W) | Cumulative Production |
|---|---|---|
| 1976 | $76 | <1 MW |
| 1990 | $8 | ~50 MW |
| 2000 | $4 | ~1 GW |
| 2010 | $1.50 | ~40 GW |
| 2020 | $0.20 | ~700 GW |
| 2024 | $0.10-0.15 | ~1,600 GW |
This 99%+ cost decline is the single most important fact in energy economics. The learning rate of ~24% (prices fall 24% for each doubling of cumulative production) has held remarkably steady for nearly 50 years. Chapters 7 and 8 explore the history and mechanisms behind this revolution.
6.5.6 PV Energy Payback
Manufacturing a solar panel requires energy: purifying silicon, growing crystals, slicing wafers, processing cells, and assembling modules. The total embodied energy is roughly 400 kWh/m2. At 20% efficiency and 20% capacity factor:
\[\text{Annual output} = 1{,}000 \text{ W/m}^2 \times 0.20 \times 0.20 \times 8{,}760 \text{ h} = 350 \text{ kWh/m}^2\text{/year}\]
Energy payback time: 400 / 350 \(\approx\) 1.1 years. Over a 25-year lifetime, a panel produces 15-25 times the energy used to make it (EROI of 15-25). This payback time has fallen from 4+ years in the 2000s as manufacturing has become more efficient.
6.6 Solar Thermal Conversion
6.6.1 Concentrating Solar Power (CSP)
An alternative to photovoltaics: concentrate sunlight to produce heat, then use that heat in a conventional thermal cycle. CSP plants are heat engines that happen to use sunlight as the heat source.
The Carnot limit applies: \[\eta_{Carnot} = 1 - \frac{T_{cold}}{T_{hot}}\]
For CSP operating at 400°C (673 K) with ambient at 40°C (313 K): \[\eta_{Carnot} = 1 - \frac{313}{673} = 53\%\]
But this is only the thermal-to-electric conversion. We must also account for:
- Optical efficiency (reflector losses, spillage, cosine effects): ~60-70%
- Receiver thermal losses (radiation, convection): ~80-90%
Overall solar-to-electric efficiency for CSP: 15-25%
The loss cascade is instructive. Starting from 1,000 W/m2 peak irradiance:
| Stage | Factor | Remaining (W/m2) |
|---|---|---|
| Latitude/time averaging | ×0.50 | 500 |
| Land use (spacing) | ×0.20 | 100 |
| Mirror reflectivity | ×0.94 | 94 |
| Receiver efficiency | ×0.85 | 80 |
| Rankine cycle | ×0.33 | 26 |
| Parasitic loads | ×0.92 | 24 |
The result: roughly 24 W/m2 average power density, similar to PV. Every CSP plant faces this cascade; the engineering challenge is minimizing each loss factor.
CSP technologies include:
Parabolic trough: Linear mirrors focus sunlight on a tube carrying heat-transfer fluid. Mature technology, operating since the 1980s in California. Temperatures: 300-400°C.
Power tower: Field of flat mirrors (heliostats) focus sunlight on a central receiver atop a tower. Higher temperatures possible: 500-600°C, improving thermal efficiency.
Dish-Stirling: Parabolic dish focuses sunlight on a Stirling engine at the focal point. Highest efficiency (~30%) but small scale and limited deployment.
6.6.2 Thermal Storage: CSP’s Advantage
CSP’s unique advantage over PV: heat is easier to store than electricity. Molten salt (typically nitrate mixtures) can store thermal energy for hours:
- Heat capacity: ~1.5 kJ/kg·K
- Operating temperature: 290-565°C
- Storage density: ~250 MJ/m3
A CSP plant can store excess daytime heat and continue generating electricity after sunset. The Gemasolar plant in Spain demonstrated 24-hour operation. The Crescent Dunes plant in Nevada stores enough heat for 10 hours of full output.
The Crescent Dunes plant in Nevada designed for 110 MW with 10 hours of storage (1.1 GWh thermal). At 40% thermal-to-electric efficiency, this provides ~440 MWh of electricity after sunset.
The molten salt (60% NaNO3, 40% KNO3, melting point ~220°C, stable to ~600°C) has a heat capacity of ~1.5 kJ/(kg·K). Operating between 290°C and 565°C (ΔT = 275 K):
\[m = \frac{1.1 \times 10^{9} \text{ Wh} \times 3{,}600 \text{ J/Wh}}{1{,}500 \text{ J/(kg·K)} \times 275 \text{ K}} \approx 9{,}600 \text{ tonnes per tank}\]
That is a lot of salt, in tanks the size of oil storage tanks. Crescent Dunes ultimately failed: a hot salt leak in 2016 caused extended downtime, the project went bankrupt in 2020, and was decommissioned in 2024. The lesson: CSP’s thermodynamic elegance does not guarantee commercial success.
6.6.3 CSP in the Real World
Real-world CSP plants illustrate both the potential and the challenges:
PS10/PS20 (Seville, Spain): The first commercial power towers. PS10 (11 MW, 624 heliostats, 115 m tower) and PS20 (20 MW, 1,255 heliostats, 165 m tower). Modest in scale but proved the concept.
Ivanpah (California): 392 MW, 173,500 heliostats, 14.2 km2. Achieved 27.6 W/m2 peak power density but only ~19% capacity factor. Generated 650-700 GWh/year against a 940 GWh target. Notable controversy: bird mortality from concentrated solar flux (~6,000/year), though for context, US cats kill 2.4 billion birds annually, windows kill 600 million, and wind turbines about 500,000.
Noor Ouarzazate (Morocco): The world’s largest concentrated solar complex at 510 MW CSP plus 72 MW PV across 25 km2, built in four phases. Demonstrates CSP at national scale in a developing country.
6.6.4 CSP Economics: The Market Verdict
The market has largely chosen PV over CSP:
| Technology | LCOE ($/MWh) |
|---|---|
| Utility PV (no storage) | 28-35 |
| PV + 4h battery | 50-70 |
| CSP (no storage) | 100-120 |
| CSP (6h storage) | 90-110 |
CSP retains a niche where long-duration dispatchable solar is valued, but PV plus batteries has become the default solar choice.
This “dispatchability” distinguishes CSP from PV. But the advantage has narrowed as battery costs fall, and batteries can store electricity from PV just as molten salt stores heat from CSP.
CSP plants have higher capital costs and lower efficiency than PV, but offer built-in storage (Security/reliability advantage). They require direct sunlight, limiting deployment to sunny, dry climates (Sustainability concern: desert ecosystems). Large CSP projects involve major construction with concentrated employment (Equity: local jobs vs. land use conflicts).
PV panels can deploy anywhere, at any scale, with minimal site preparation. Costs have plummeted (improving Equity through affordability). But PV requires external storage or backup for reliability.
The market has largely chosen PV—costs fell faster, deployment is simpler, and batteries provide an alternative storage solution. CSP retains a niche where thermal storage is particularly valuable, but the technology has struggled to compete.
6.7 Power Density Revisited
Returning to a theme from Chapter 1: how much land does solar energy require?
6.7.1 Theoretical Maximum
If we could convert 100% of incident solar energy: \[\text{Power density} = 1000 \text{ W/m}^2 \times \text{capacity factor}\]
With a 25% capacity factor (typical for fixed-tilt PV in good locations): \[\text{Maximum} = 1000 \times 0.25 = 250 \text{ W/m}^2\]
But we can’t convert at 100% efficiency.
6.7.2 Real-World Power Density
A solar farm includes:
- Panels: ~20% efficient
- Spacing: Panels can’t be touching; allow ~2× ground area for spacing, access, inverters
- Capacity factor: ~25-30% in sunny locations
Typical utility-scale solar: 5-8 W/m2 of land area
This is roughly 20× lower than the theoretical maximum—a combination of conversion efficiency, spacing requirements, and temporal variation.
U.S. electricity consumption: ~4,000 TWh/year Average power: 4,000 TWh / 8,760 h = 450 GW
If solar provided 100% (ignoring storage issues):
- At 6 W/m2: 450 GW / 6 W/m2 = 75 billion m2 = 75,000 km2
That’s a square about 270 km × 270 km—roughly the size of South Carolina. Or ~1% of the continental U.S. land area.
Is this a lot? Compared to the land area already devoted to roads, parking lots, and buildings, not really. But compared to what we’re used to for energy infrastructure (power plants are small; the fuel comes from elsewhere), it represents a fundamental shift in land use.
6.7.3 Smil’s Critique
Vaclav Smil has argued that solar’s low power density represents a fundamental constraint on energy transition. His point: modern civilization was built on fossil fuels precisely because their high power density (extraction from compact sources, conversion in compact plants) enabled concentrated industrialization.
Returning to distributed, low-power-density sources means either: 1. Covering vast areas with energy infrastructure 2. Accepting lower total energy consumption 3. Making dramatic improvements in energy efficiency
There’s truth in this critique. But it shouldn’t be overstated. The land required for solar is large compared to conventional power plants but small compared to the land we already use for agriculture, urbanization, and transportation. The question is not whether solar can physically provide our energy needs—it can—but whether the transition can occur at the pace required for climate stabilization.
6.8 Key Concepts
- The two paths: Fossil fuels take the “long path” (photosynthesis → geological storage); PV takes the “short path” (direct conversion)
- Solar constant: 1,361 W/m2 at Earth’s orbital distance (derived in Chapter 4)
- Air Mass (AM): Path length through atmosphere; AM1.5 (1,000 W/m2) is the standard test condition
- Band gap: Energy difference between valence and conduction bands; determines which photons are absorbed
- Shockley-Queisser limit: ~33% maximum efficiency for single-junction cells (physics, not engineering)
- PV vs. photosynthesis: PV achieves 20-25% efficiency vs. 0.5-2% for plants (10-40× advantage)
- Multi-junction cells: Stack different band gaps to capture more of the spectrum
- CSP vs. PV: Heat engines vs. direct photovoltaic conversion; CSP offers storage, PV offers lower cost
- Power density: Real-world solar farms achieve 5-20 W/m2 of land area (10-40× higher than biomass)
6.9 Exercises
Photon flux: The AM1.5 solar spectrum delivers about 4.3×1017 photons/s/cm2 with energy above silicon’s band gap (1.12 eV). If each absorbed photon creates one electron-hole pair, what is the maximum possible current density (A/cm2)? How does this compare to typical silicon cell short-circuit current of ~40 mA/cm2?
Thermalization losses: A photon with λ = 400 nm (violet) has energy E = hc/λ = 3.1 eV. When absorbed by silicon (Eg = 1.12 eV), what fraction of the photon’s energy becomes heat?
Land requirement comparison: Compare the land area needed to generate 1 GW (average) from: (a) solar PV at 6 W/m2, (b) wind at 2 W/m2, (c) biomass at 0.5 W/m2. Express your answers in km2 and as a fraction of Rhode Island’s area (4,000 km2).
Latitude effects: Calculate the annual average solar irradiance on a horizontal surface at: (a) equator, (b) 30°N, (c) 60°N. Assume clear skies and account for the cosine of solar zenith angle averaged over the day and year. (Hint: The declination angle varies from +23.4° to -23.4° over the year.)
CSP vs. PV economics: A CSP plant costs $4,000/kW with 40% capacity factor (using 10-hour thermal storage). A PV plant costs $1,000/kW with 25% capacity factor. Calculate the cost per kWh over a 25-year lifetime (ignore O&M and financing for simplicity). At what battery cost does PV + 4-hour storage match the value proposition of CSP?
Theoretical multi-junction: Design an ideal three-junction solar cell by selecting band gaps that minimize combined transmission and thermalization losses. The solar spectrum can be approximated as having equal energy in ranges 280-700 nm, 700-1400 nm, and 1400-2500 nm. What efficiency might you expect compared to the single-junction Shockley-Queisser limit?
This chapter establishes the Principles of direct solar energy conversion—the physics that sets hard limits on what is achievable. The solar constant (from Chapter 4), the band gap, and the Shockley-Queisser limit are fixed by physics, not engineering.
Comparing the two paths:
- Long path (Chapter 4): Sun → photosynthesis (0.5-2%) → burial (0.1%) → fossil fuel → combustion
- Short path (this chapter): Sun → PV (20-25%) → electricity
The short path is fundamentally more efficient but lacks the geological storage that makes fossil fuels dispatchable. This tradeoff (efficiency vs. storability) drives much of the energy transition challenge.
The next two chapters will show how Principles enable Technologies (crystalline silicon, thin films) that became Products (commercial panels) through a remarkable cost-reduction story. Chapters 8-9 trace the Policy decisions that shaped deployment.