Browser-based power budget tool for CubeSats from 1U up to 12U. Adjust body-mounted and deployable solar panels, cell efficiency, orbit altitude, eclipse fraction, payload and bus loads, and immediately see orbit period, orbit-average generation, power margin and required battery capacity for early sizing trades.
Parameters
CubeSat size
Pick a form factor (10 cm cube unit)
Cell efficiency η
%
Triple-junction GaAs is typically 28-30%
Body-mounted panels
faces
Number of solar-cell faces bonded to the body
Deployable panel area
cm²
Total area of sun-tracking deployable wings
Orbit altitude
km
ISS sits at 400 km; SSO at 500-800 km
Eclipse fraction (per orbit)
Share of the orbit spent in Earth shadow
Payload power
W
Mission instrument and radio average draw
Bus baseline power
W
OBC, sensors, ADCS housekeeping power
Battery efficiency
%
Round-trip charge/discharge (Li-ion 88-92%)
Results
—
Orbit period (min)
—
Body panel power (W)
—
Deployable power (W)
—
Orbit-average power (W)
—
Power margin (%)
—
Required battery (Wh)
—
Orbit, Sun and CubeSat schematic
Sun rays (yellow) hit body-mounted and deployable panels. Inside Earth's shadow the spacecraft runs on its battery, whose level (green to red) is shown by the bar below.
P_gen is the gross solar generation (including cosine effects), η_eclipse the eclipse fraction (typically 0.30-0.40 in LEO), t_eclipse the eclipse duration.
Body-mounted panels see an average illumination factor 1/π ≈ 0.318 over rotation. S = 1361 W/m² (solar constant), R_E = 6378.137 km, h = orbit altitude.
CubeSat Power Budget Design
🙋
I keep reading about university CubeSat launches. Can a 10 cm cube really do useful science? It seems impossible to fit enough solar cells on something that small.
🎓
That's exactly why the first thing every CubeSat team does is the power budget. A 1U (10×10×10 cm, ~1.33 kg) has to host attitude control, comms and a science payload, all fed by a few square centimetres of solar cells on its outer faces. So you list every consumer, watt by watt, and check the orbit-average balance. If that comes out negative, no clever radio design can save you.
🙋
With the defaults on the left (3U, 29% cells, four faces), the body panels make about 12 W, but the orbit-average drops to 7.8 W. Where do the missing watts go?
🎓
Two effects. First, geometry: a tumbling cube only ever points half its faces at the Sun, and the rotation-averaged projection factor works out to 1/π ≈ 0.318. Second, eclipse: in LEO about 35% of every 90-minute orbit is spent in Earth's shadow, so 12 W × (1 − 0.35) ≈ 7.8 W. Deployable wings that track the Sun keep about 0.9 of full illumination, so they roughly double the W/cm² of body-mounted cells.
🙋
What about during eclipse — does everything just stop?
🎓
That's what the battery is for. You charge during sunlight and discharge through eclipse: E_batt = P_load · t_eclipse / η_batt. With the defaults the eclipse lasts ~0.54 h (32 min), so 3.5 W × 0.54 / 0.88 ≈ 2.14 Wh, or ~290 mAh at 7.4 V. In real hardware you keep the depth-of-discharge to 30-50% for cycle life, so you actually install 600-1000 mAh.
🙋
How much margin should I aim for? The verdict says "OK above 30%", but the defaults show 123%. That has to be overkill, right?
🎓
Good catch. The defaults model a very lean baseline — four-face 3U with a 3.5 W load. Real CubeSats add a transmit amplifier (8-10 W when keying), maybe an electrothermal thruster (5-20 W), and instruments that all want to be on at the same time. The NASA and ESA design rules ask for 30% EOL margin at PDR and 20% at CDR. Since this tool prints BOL ideal numbers, factor in radiation, attitude error and battery losses by treating 30-50% as your floor here. Too much margin is also a sign — it tells you that you can drop a face, save mass, or fly a more ambitious instrument.
🙋
What are the classic failures when real university CubeSats run short on power?
🎓
Three patterns dominate. One, attitude control never quite settles and the satellite spends a long time off-Sun — half the expected generation. Two, the Li-ion pack gets cold (below −20 °C) and loses most of its usable capacity. Three, on-orbit radiation pushes BOL toward EOL faster than expected. So designers apply chained derating: 0.7 for pointing, 0.9 for thermal, 0.85 for radiation. Even cutting-edge programmes like SpaceX Starlink, JAXA Hayabusa2 or NASA MarCO simulate the worst case — everything on, battery cold, one panel half-shadowed — before they freeze the EPS design.
FAQ
First compute the gross solar array power P_gen = A·S·η·k. A is the array area (m²), S is the on-orbit solar constant (~1361 W/m²), η is the cell efficiency, and k is the illumination factor. For body-mounted panels on a tumbling satellite the rotation-averaged factor is k≈1/π=0.318, while sun-tracking deployable wings use k≈0.9. Then subtract the eclipse fraction with P_avg = P_gen·(1 − η_eclipse). η_eclipse is typically 0.30 to 0.40 in LEO.
The minimum requirement is to feed the full load through the longest eclipse: E_batt = (P_load · t_eclipse) / η_batt. η_batt is the round-trip efficiency, typically 88-92% for Li-ion. In practice you also limit Depth of Discharge (DOD) to 30-50%, so the installed cell capacity is 2 to 3 times the calculated minimum. This tool reports the result in Wh and as mAh assuming a 7.4 V two-cell Li-ion pack.
Surface area roughly scales with the U count. One side of a 1U (10×10×10 cm) is 80 cm², while a long side of a 3U (10×10×30 cm) is 240 cm². With four 29% efficient body-mounted panels you can expect roughly 4 W orbit-average on a 1U and around 12 W on a 3U. On a 1U typically only 0.5 to 1 W is left for the payload, so any meaningful radio or propulsion module pushes you to deployable wings or to 3U and larger.
NASA and ESA design guidelines recommend at least 30% EOL (End-of-Life) margin at PDR and 20% at CDR. The numbers this tool prints are BOL (Beginning-of-Life) ideal values, so once you add radiation degradation and attitude error you want at least 30-50% margin here to be realistic. Margin < 0% is a hard NO-GO; < 30% is shown as a warning.
Real-world applications
University and startup demonstrators: from Tokyo University's CUTE-1 (1U, 2003) onward, dozens of universities — Kyutech, Tokyo Tech, Hokkaido and many others — have flown their own CubeSats. Commercial constellations such as Planet Labs (Dove 3U) and Spire (Lemur 3U) build Earth-imaging and GNSS radio-occultation fleets out of the same form factor. They all start with a power-budget trade exactly like the one this tool does.
Scientific deep-space missions (NASA MarCO, ESA M-ARGO): NASA's MarCO (2018) was the first CubeSat to reach Mars, relaying the InSight landing. At Mars the solar constant drops to less than half of Earth's, forcing very large deployable wings and aggressive duty cycling — power budgets at planetary distance are several times tighter than in LEO.
Commercial constellations (Starlink, OneWeb): Starlink V2 mini is larger than a CubeSat, but its mass-production mindset is a direct descendant of CubeSat design. Whether the spacecraft is 1.33 kg or 800 kg, balancing solar array efficiency, eclipse loss and battery cycle life decides the operating margin and ROI of the constellation.
Concept studies for larger spacecraft: even flagship missions like Hayabusa2 or SLIM start with quick power-budget spreadsheets to lock the order of magnitude of arrays and batteries. Detailed design migrates to STK or SOAP for attitude-coupled simulation, but a tool like this one decides in the first hours whether a concept closes at all.
Common pitfalls
The biggest trap is to size the EPS only at BOL (Beginning-of-Life). The numbers here are ideal. In orbit the cells lose 10-20% over five years from radiation, and atomic oxygen and UV haze the cover glass on top. Most teams assume EOL = 0.7-0.8 × BOL. A design that shows 30% margin in this tool can sit at zero by end of life, so always re-evaluate at EOL = 0.7-0.8 × BOL.
The second trap is to assume perfect attitude control. The 1/π illumination factor is the long-term average over a fully tumbling cube. Real attitude errors and post-deployment transients drop it well below that, sometimes to zero for minutes at a time. The first days or weeks after deployment are usually attitude-unstable, so the EPS has to survive on the battery alone — verify that the battery does not run dry in those scenarios.
Finally, battery capacity alone does not guarantee eclipse survival. Li-ion cells lose more than half their usable capacity below −20 °C and age fast above +50 °C. CubeSat on-orbit temperatures swing between −40 °C and +60 °C, so the pack needs heaters (a few W of survival heating) inside the budget. Keep the depth-of-discharge to 30-50% for multi-year cycle life, which means installing 2-3× the calculated capacity.
How to Use
Set body-mounted solar panel efficiency (12–22% for typical multi-junction cells) and quantity (1–6 panels for 1U to 12U CubeSats)
Enter deployable panel area in cm² and select orbit altitude (400–600 km for LEO, affects eclipse duration)
Simulator calculates orbit period, eclipse ratio, and generates orbit-average power; compare power margin (%) against mission payload draw (typically 5–15 W for imaging or comms)
Worked Example
3U CubeSat in 500 km sun-synchronous orbit: 4 body-mounted panels (20% efficiency, ~10×10 cm each = 400 cm²) + 200 cm² deployable panel. Solar constant 1361 W/m² at orbital edge, eclipse duration ~35 min per 94 min orbit. Body panels generate ~6.5 W average, deployables add ~2.8 W. Total orbit-average power 9.3 W. With 8 W payload load, margin = 16%. Li-ion battery (2000 mAh, 7.4 V) supplies 14.8 Wh eclipse reserve.
Practical Notes
Body-mounted panels face trade-off: more area reduces deployable real estate; 4–6 panels typical for 3U imaging missions
Deployable panels suffer 8–12% power loss during deployment cycling; account in efficiency slider
Equatorial vs. polar orbit changes eclipse duration by ±8 min; rerun at mission inclination for accuracy
Account for 15% degradation over 2-year mission lifetime when setting margin threshold