How to Do Real Energy Research at Home: A High School Student’s Guide to Free Tools, Affordable Kits, and Public Databases

Energy research used to mean a university lab, a shared cleanroom, and a faculty advisor with grant money. That world has cracked wide open. You can now build a working solar cell on your kitchen counter, model a lithium battery on a free cloud GPU, and pull live grid and weather data from public APIs, all from a bedroom desk.

This guide is your starting point. It covers three things: the affordable kit you need, the free software professionals actually use, and the public databases that count as real data. Together, they let a high school student run a project that holds up at a regional fair or ISEF.

Why This Is Possible Now

The first shift is in components. Small solar cells, Peltier modules, 18650 cells, piezo films, Arduino boards, and USB potentiostats now cost a few dollars each. A complete tabletop energy lab fits in a shoebox.

The second shift is in software. Battery simulators (PyBaMM), circuit simulators (SPICE), CFD solvers (OpenFOAM), solar yield models (pvlib), and DFT engines (Quantum ESPRESSO) are all free and run on a laptop or a Google Colab GPU. Tools that cost universities six figures in the 2000s are now an install away.

The third shift is in data. NASA POWER gives you decades of irradiance and weather. NOAA and METAR feeds give you wind. PurpleAir gives you air-quality data. The Materials Project, CoRE-MOF, and NASA Ames Battery datasets give you ready-to-analyze experimental records. The U.S. Energy Information Administration publishes the grid itself.

Put those three together. A kitchen counter, a $50 kit, and a laptop can now run a real energy experiment, simulate it, and compare it against the same datasets professional researchers use.

The Energy Home Kit

Group your purchases by what they let you measure, build, and test.

Power generation parts

• Small monocrystalline PV cells (1V to 6V, around $2 to $10 each)
• Peltier thermoelectric modules (TEC1-12706, around $5)
• Hobby DC motors used as micro generators ($3 to $8)
• Piezoelectric discs and PVDF film strips ($1 to $10)
• PTFE tape, aluminum foil, and PDMS sheets for triboelectric builds

Electrochemistry and storage parts

• Graphite rods (from pencil leads or salvaged batteries)
• Steel wool, copper sheet, zinc-galvanized nails, tin foil
• Salvaged 18650 cells from old laptop packs
• Activated carbon, baking soda, NaCl, NaOH (drain cleaner), KMnO4
• Agar powder, gelatin, corn starch for bio-based separators

Sensors and measurement

• Digital multimeter ($15 to $30)
• USB potentiostat or LCR meter ($30 to $80)
• DHT22 / BMP280 temperature and humidity sensors
• TSL2591 or phone-based lux meter for irradiance
• Cheap pH and conductivity probes (USB or analog)
• Thermal camera phone attachment (FLIR ONE or InfiRay, optional)

Electronics and control

• Arduino Uno or ESP32 board ($10 to $20)
• Breadboard, jumper wires, resistors, capacitors, op-amps
• 3D-printed turbine blades, ducts, and enclosures (school maker space or library)
• Box fan as a tabletop wind tunnel
• Fish tank as a wave/flow flume

Bio and biochar prep

• Spent coffee grounds, banana peels, citrus peels (carbon precursors)
• Mason jars and a household toaster oven for low-oxygen pyrolysis (with adult supervision)
• Chlorella culture starter kit, mushroom spawn, kitchen compost

A full starter kit, sourced over a few weeks, lands in the $80 to $200 range. You will not need all of it for any one project. Buy in waves as your question sharpens.

Signature Technique: Reading Energy With an Arduino and Your Phone

The single technique that unlocks the most Energy projects is electrical characterization with a microcontroller plus phone-based sensing. Almost every phenomenon in this guide ends in "measure voltage, current, or power across a condition". An Arduino plus a phone gives you that, with the precision you need.

Here is the 5-step workflow.

1. Wire your device under test (DUT) to an Arduino analog input through a known load resistor. For higher voltages, add a simple voltage divider. For higher currents, add a sense resistor and an op-amp.
2. Sample at a fixed rate (10 Hz to 1 kHz depending on the phenomenon) and stream readings over USB serial.
3. Log the stream with the Arduino IDE Serial Plotter, with PySerial in Python, or directly into a Google Sheet via a simple script.
4. Pair each electrical sample with an environmental measurement from your phone: lux, temperature, sound dB, or IMU acceleration. Timestamp both streams.
5. Compute the derived quantity that matters: power density, Faradaic efficiency, round-trip efficiency, power coefficient, capacity fade. Plot it against your independent variable.

That loop, repeated cleanly across one good independent variable and three or more replicates per level, is the backbone of a competition-grade Energy project.

The Dry-Lab Side: Free Software You Can Install Today

Group the tools by what they let you do.

Circuit and device simulation

• LTspice — free SPICE simulator for solar bypass diodes, battery cell models, and TENG charge transfer circuits.
• Ngspice / PySpice — scripted SPICE runs you can sweep from Python.
• PV Lighthouse and pvlib — pvlib is a Python library for full PV system yield modeling at any latitude.

Batteries and electrochemistry

• PyBaMM — open-source lithium-ion battery simulator. Runs degradation, fast-charge, and duty-cycle studies.
• CasADi + PyBaMM — for fitting model parameters to your own measured cycles.

Fluids, heat, and structures

• OpenFOAM — professional CFD. Use it for wind, water, ventilation, and heat-exchanger studies.
• SU2 — easier-to-learn open-source CFD alternative.
• FreeCAD and OpenSCAD — parametric CAD for turbine blades, ducts, and enclosures.
• FEniCS / SolidsPy — finite-element solvers for structural and topology optimization.
• EnergyPlus — the building-energy simulator used by professional HVAC engineers.

Materials and chemistry

• Quantum ESPRESSO on Google Colab — DFT calculations for perovskites, catalysts, and 2D materials, on a free cloud GPU.
• RASPA — molecular simulation of gas adsorption in MOFs, useful for hydrogen storage screening.
• COBRApy — flux-balance analysis for engineered microbes (biohydrogen, biogas).

Process and life-cycle modeling

• CoolProp + Python — thermodynamic properties for any working fluid in your cycle.
• openLCA with the free ecoinvent starter — life-cycle analysis that holds up at a fair.

Machine learning

• scikit-learn, PyTorch, and Google Colab — for forecasting wind, predicting battery state of health, and screening catalysts.

Running the same tools as a working engineer changes how the project feels. You stop reproducing demos and start producing results.

Public Databases That Count as Real Data

Group your data sources by what they measure.

Weather, solar, and wind

• NASA POWER — decades of hourly irradiance, temperature, and wind at any latitude/longitude.
• NOAA and METAR — historical wind speed, direction, and pressure for any U.S. airport.
• NREL National Solar Radiation Database (NSRDB) — high-resolution U.S. solar resource data.
• PurpleAir — crowdsourced PM2.5 and AQI feeds, useful for soiling and air-quality work.

Buildings, grid, and load

• U.S. EIA Open Data — electricity generation, prices, and demand by region.
• OpenEI — energy datasets, residential load profiles, and tariff data.
• Pecan Street and UMass Smart Datasets — anonymized household smart-meter time series.
• Google Project Sunroof — rooftop solar potential by address.

Batteries and materials

• NASA Ames Prognostics Battery Dataset — cycled 18650 data, the standard for SOH and RUL work.
• Sandia and CALCE battery datasets — additional public cycling data.
• Materials Project — DFT-computed properties for hundreds of thousands of inorganic materials.
• CoRE-MOF and QMOF — curated MOF databases for hydrogen and CO2 storage screening.

Bio and waste streams

• USDA Waste-Stream and Agricultural Residue datasets — feedstock composition for biogas modeling.
• NCBI 16S rRNA databases — for pairing with mail-in microbial sequencing of fuel cell consortia.

Demographics and equity layers

• U.S. Census and ACS — pair with energy data for equity-aware solar siting.
• EJScreen — environmental-justice indicators for policy-flavored projects.

Re-analyzing a public dataset is a real research path. A clean, well-framed analysis of NASA POWER or NASA Ames data can stand on its own at a fair.

How to Combine Wet and Dry: The Strongest Project Shape

Pattern A: Build, measure, then model. You construct a small physical prototype (a DSSC, a TEG, a Savonius turbine, a TENG insole) and characterize it with your Arduino rig. Then you build a simulation in pvlib, SPICE, OpenFOAM, or PyBaMM that predicts the same numbers, and compare. The gap between model and measurement is your discussion section.

Pattern B: Screen in silico, then validate one case. You run a sweep in software first (a DFT screen of perovskite absorbers, a CFD sweep of duct geometries, a PyBaMM sweep of duty cycles) and pick the predicted winner. You then build only that one physical case and confirm or refute the prediction.

Both patterns work because judges reward students who can reason across scales. A measurement alone is a demo. A measurement plus a matching model is a research project.

Choosing a Phenomenon That Has Not Been Done

Originality is a process, not a guess.

1. Search Google Scholar for your phenomenon plus the words "high school" or "low cost". Read the abstracts of the most-cited papers from the last five years. Note what they varied and what they did not.
2. Search the Society for Science abstracts archive for past ISEF and Regeneron projects in Energy: Sustainable Materials and Design. Filter by your subcategory. Read the abstracts, not the titles.
3. Search PubMed and the IEEE Xplore database for the underlying mechanism (electrolysis catalysts, TENG materials, battery aging models). Look for review articles. Reviews tell you what is open.

If you find ten papers on your exact phenomenon, change one variable: the material, the geometry, the dataset, the scale, or the comparison. Finding adjacent prior work is good news, because it tells you the question is real and that your variation can be the contribution.

A Realistic Timeline

• One to two weeks: A focused replication or measurement. Build one DSSC, characterize five voltage curves, write it up.
• One to two months: A full hybrid project for a regional fair. Pattern A or B above, three to five conditions, replicates, error bars, and a simulation alongside.
• Full year: An ISEF-track project. Multiple phases, an iterated prototype, real statistics, and a clear novelty claim grounded in your literature review.

If this is your first research project, start with the one to two week version. Finishing a small project teaches you more than starting a big one.

A Starter Checklist

1. A clean, well-lit workspace with a power strip, ventilation, and a place to store wet samples safely.
2. A free Google Colab account, signed in and tested with one GPU notebook.
3. A local Python environment (Anaconda or Miniconda) with NumPy, pandas, matplotlib, scikit-learn, and pvlib installed.
4. One simulator from the dry-lab list installed and verified on a sample input (LTspice, PyBaMM, OpenFOAM, or Quantum ESPRESSO on Colab).
5. An Arduino IDE setup with a working "blink" and a working analog-read sketch streaming to your laptop.
6. A bound lab notebook or a dated digital notebook (Notion, OneNote, or a Git repo) where every measurement gets logged with date and conditions.
7. A one-line research question written at the top of page one: "How does X affect Y in Z?"

When those seven items are in place, you are ready to pick a phenomenon.

Where to Go Next

ISEF's Energy: Sustainable Materials and Design category has seven subcategories. Each one fits the kit and software on this page.

• Biological Process and Design (BIO): Microbial fuel cells, algae biodiesel, anaerobic digestion, biochar electrodes, and engineered microbes for biohydrogen.
• Solar Process, Materials, and Design (SOL): Dye-sensitized cells, perovskite screening, trackers, bifacial layouts, soiling, and luminescent concentrators.
• Energy Storage (EST): Lithium-ion aging, supercapacitors from biomass, gravity and thermal batteries, novel separators, and state-of-health estimation.
• Wind and Water Movement Power Generation (FLD): Savonius and vertical-axis turbines, biomimetic blades, vortex harvesters, wave and tidal converters, and wind forecasting.
• Hydrogen Generation and Storage (HYD): Earth-abundant electrocatalysts, photoelectrochemical splitting, MOF storage, LOHC modeling, and gasochromic leak sensors.
• Thermal Generation and Design (THR): Thermoelectrics, geothermal sizing, phase-change and salt-hydrate storage, passive ventilation, and pyroelectric harvesting.
• Triboelectricity and Electrolysis (ELC): TENGs in shoes, rain, masks, and waves, biodegradable triboelectric films, and electrolytic CO2 reduction.
• Other (OTH): Whole-home energy twins, RF and acoustic harvesting, microgrid RL controllers, life-cycle analysis, and equity-aware solar siting.

Each subcategory has its own MehtA+ project guide that builds directly on the kit, software, and datasets above. Pick the one that pulls you in most, and start there. The lab door is already open.

Project ideas in this category (87)