1619  Energy Harvesting Practical Guide

Real-World Examples, Common Pitfalls, and Advanced Techniques

1619.1 Learning Objectives

TipWhat You’ll Learn

After completing this chapter, you will be able to:

  • Apply energy harvesting to real-world IoT applications
  • Avoid common design mistakes and pitfalls
  • Implement Maximum Power Point Tracking (MPPT)
  • Design hybrid harvesting systems for improved reliability
  • Select appropriate storage technologies

1619.2 Common Pitfalls

WarningAvoid These Design Mistakes
  1. Underestimating variability - Solar and vibration sources vary significantly; design for worst case
  2. Ignoring startup energy - Systems need minimum energy to boot; plan for cold-start scenarios
  3. Forgetting leakage currents - Supercapacitors self-discharge; account for this in storage sizing
  4. Over-optimizing for peak - Design for average conditions with adequate margin
  5. Neglecting aging - Battery capacity degrades 20-30% over lifetime; plan accordingly
  6. Single-source dependency - Hybrid systems provide redundancy and higher reliability

1619.2.1 Detailed Pitfall Analysis

1619.2.1.1 Pitfall 1: Underestimating Variability

Problem: Designing for average solar irradiance ignores cloudy days, winter months, and shading.

Solution: Use the worst-month design approach:

  • For solar: Design for December irradiance (Northern Hemisphere)
  • For vibration: Design for minimum expected machine operating hours
  • Add 30-50% safety margin to all calculations
Example:
- Average solar harvest: 50 mWh/day
- Worst month factor: 0.3 (winter)
- Design harvest: 50 × 0.3 = 15 mWh/day
- With 30% margin: 15 / 0.7 = 21 mWh/day available

1619.2.1.2 Pitfall 2: Ignoring Startup Energy

Problem: The device needs energy to boot, but storage is empty after long periods without harvesting.

Solution: Implement staged startup:

  1. Accumulate energy until minimum threshold (e.g., 3.0V)
  2. Enable boost converter only when threshold reached
  3. Use ultra-low-quiescent PMIC (< 1 µA when disabled)
  4. Include “power good” signal to MCU

1619.2.1.3 Pitfall 3: Forgetting Leakage Currents

Problem: Supercapacitor self-discharge can drain storage overnight.

Typical self-discharge rates:

Storage Type Self-Discharge Rate
Supercapacitor (EDL) 10-30% per day
Supercapacitor (hybrid) 5-10% per day
LiPo Battery 2-3% per month
LiFePO4 Battery 1-2% per month

Solution: For systems with overnight non-harvesting periods:

  • Use hybrid storage (supercap for peaks, battery for overnight)
  • Size supercap for daily cycle only
  • Account for self-discharge in energy budget

1619.3 Practical Examples

1619.3.1 Example 1: Solar-Powered Environmental Sensor

Requirements:

  • Measure temperature, humidity every 5 minutes
  • Transmit via LoRaWAN once per hour
  • Deploy in outdoor location

Power Profile Analysis:

Mode Current Duration Energy
Deep Sleep 2 µA 59 min/hr 0.39 mWh
Sensor Read 5 mA 50 ms × 12/hr 0.01 mWh
LoRa TX 100 mA 100 ms/hr 0.03 mWh
Total/Hour - - 0.43 mWh
Total/Day - - ~10 mWh

Solution Design:

  • Solar panel: 25 cm² @ 18% efficiency = ~40 mW peak
  • Available sunlight: 6 hours (worst case) × 40 mW = 240 mWh
  • System efficiency: 70%
  • Usable energy: 240 × 0.7 = 168 mWh/day
  • Margin: 168 / 10 = 16.8x energy margin

Storage Sizing (5-day autonomy):

  • Energy needed: 10 mWh × 5 days = 50 mWh
  • With 30% margin: 65 mWh
  • 1000 mAh LiPo @ 3.7V = 3700 mWh capacity
  • Result: 70+ days autonomy on full charge

1619.3.2 Example 2: Industrial Vibration Monitor

Requirements:

  • Monitor vibration continuously
  • Alert on anomaly detection
  • No battery replacement in 10 years

Environment Assessment:

  • Machinery: 120 Hz rotation, 0.8g acceleration
  • Operating hours: 16 hours/day (two shifts)
  • Available thermal differential: 15°C on motor housing

Power Profile:

Mode Power Duration Daily Energy
Continuous monitoring 0.5 mW 24 hr 12 mWh
Alert transmission 100 mW 1 sec × 10/day 0.3 mWh
Total/Day - - ~12.5 mWh

Solution Design:

Vibration harvester (primary):

  • Medium piezo harvester on 120 Hz source
  • Expected output: 2 mW × 16 hours = 32 mWh/day

Thermal harvester (backup):

  • 16 cm² TEG @ 15°C ΔT
  • Expected output: 5 mW × 16 hours = 80 mWh/day (when ΔT available)

Total available: 32-112 mWh/day vs 12.5 mWh consumed

Storage:

  • 5F supercapacitor bank for alert bursts
  • Supercap energy: 0.5 × 5 × (5² - 2.5²) / 3.6 = 13 mWh usable
  • Result: 2.5-9x margin, supercap handles burst TX without battery

1619.3.3 Example 3: Wearable Health Tracker

Requirements:

  • Continuous heart rate monitoring
  • Skin temperature measurement
  • BLE transmission every 30 seconds
  • All-day operation

Multi-Source Harvesting:

Source Configuration Expected Power
Body heat (TEG) 4 cm², 5°C ΔT 0.1 mW
Motion (kinetic) Wrist movement 0.2 mW
Indoor light 4 cm² solar 0.05 mW
Total - 0.35 mW average

Power Budget:

Mode Current @ 3V Duty Average Power
Sleep 3 µA 98% 0.009 mW
Sensing 500 µA 1% 0.015 mW
BLE TX 15 mA 1% 0.45 mW
Total Average - - ~0.5 mW

Challenge: Consumption (0.5 mW) > Harvest (0.35 mW)

Solutions:

  1. Reduce BLE TX duty cycle (every 60s instead of 30s)
  2. Add small 100 mAh battery as buffer
  3. Optimize BLE packet size (fewer bytes = less TX time)
  4. Use energy-aware transmission scheduling

1619.4 Advanced Topics

1619.4.1 Maximum Power Point Tracking (MPPT)

MPPT algorithms continuously adjust the load impedance to extract maximum power from variable sources:

\[P_{max} = \frac{V_{oc}^2}{4 \cdot R_{internal}}\]

The maximum power is transferred when load impedance matches source impedance.

Common MPPT Techniques:

Technique Complexity Efficiency Best For
Fractional Voc Low 90-95% Stable sources
Perturb & Observe Medium 95-99% General use
Incremental Conductance High 98-99% Rapidly changing
Neural Network Very High 99%+ Research

Fractional Open Circuit Voltage:

For solar cells, MPP occurs at approximately 0.76 × Voc:

1. Periodically measure Voc (disconnect load)
2. Set operating point to 0.76 × Voc
3. Repeat every few seconds

Simple and effective for slowly-varying solar inputs.

Perturb & Observe (P&O):

1. Measure current power P
2. Perturb operating voltage by small ΔV
3. Measure new power P'
4. If P' > P: continue in same direction
5. If P' < P: reverse direction
6. Repeat continuously

Works well for most applications but can oscillate around MPP.

1619.4.2 Hybrid Harvesting System Design

Architecture Options:

%%{init: {'theme': 'base', 'themeVariables': {'primaryColor': '#2C3E50', 'primaryTextColor': '#fff', 'primaryBorderColor': '#16A085', 'lineColor': '#16A085', 'secondaryColor': '#E67E22', 'tertiaryColor': '#ECF0F1'}}}%%
flowchart TD
    subgraph A["Option A: Separate PMICs"]
        A1[Solar] --> A2[PMIC 1]
        A3[Thermal] --> A4[PMIC 2]
        A2 --> A5[OR-ing]
        A4 --> A5
        A5 --> A6[Storage]
    end

    subgraph B["Option B: Multi-Input PMIC"]
        B1[Solar] --> B2[Multi-Input<br>PMIC]
        B3[Thermal] --> B2
        B2 --> B4[Storage]
    end

    subgraph C["Option C: Cascaded"]
        C1[High-V Source] --> C2[Primary PMIC]
        C3[Low-V Source] --> C4[Boost to Primary]
        C4 --> C2
        C2 --> C5[Storage]
    end

    style A6 fill:#27AE60,color:#fff
    style B4 fill:#27AE60,color:#fff
    style C5 fill:#27AE60,color:#fff

Design Considerations:

  1. Voltage matching - Sources with similar voltage ranges can share PMIC
  2. Power priority - Higher-power sources should have priority path
  3. Isolation - Prevent reverse current between sources
  4. Efficiency - Minimize conversion stages

1619.4.3 Storage Technology Comparison

Parameter Supercapacitor Li-Ion Battery LiFePO4 Solid-State
Energy Density 5-10 Wh/kg 150-250 Wh/kg 90-120 Wh/kg 200-400 Wh/kg
Cycle Life 500k-1M 500-2000 2000-5000 1000-10000
Charge Time Seconds 1-4 hours 1-2 hours Minutes
Self-Discharge High (10-30%/day) Low (2-3%/month) Very Low Very Low
Temperature Range -40 to +70°C 0 to +45°C -20 to +60°C -20 to +80°C
Best Use Case Peak power, buffering Energy storage Long life, safety Premium IoT

Selection Guidelines:

  • Supercapacitors: TX bursts, motor start-up, rapid charge cycles
  • Li-Ion: Maximum energy in small form factor, moderate temperature
  • LiFePO4: Long deployment life, outdoor temperature extremes
  • Solid-State: Premium applications, wide temperature, high reliability

1619.5 Design Guidelines Reference

1619.5.1 Solar Harvesting Guidelines

  • Outdoor: 10-100 mW/cm² power density available
  • Indoor: 10-100 µW/cm² (100-1000x less than outdoor)
  • Use MPPT for 15-30% efficiency gain
  • Consider panel orientation and shading
  • Account for seasonal variation (use worst-month design)
  • Typical panel efficiency: 15-25%

1619.5.2 Vibration Harvesting Guidelines

  • Match harvester resonance frequency to source
  • Piezoelectric: high voltage, low current output
  • Electromagnetic: low voltage, higher current output
  • Broadband harvesters for variable-frequency sources
  • Typical output: 0.1-10 mW depending on source
  • Requires mechanical coupling to vibration source

1619.5.3 Thermal Harvesting Guidelines

  • Requires sustained temperature differential
  • Body heat applications: ΔT ~ 5-15°C, ~10-50 µW/cm²
  • Industrial applications: ΔT > 20°C, ~1-10 mW/cm²
  • Use heat sinks on cold side to maximize ΔT
  • Seebeck coefficient: ~20-50 mV/K for Bi2Te3 TEGs
  • TEG efficiency: 3-8% typical

1619.5.4 RF Harvesting Guidelines

  • Very low power: typically < 100 µW from ambient
  • Best for supplementary power or dedicated TX applications
  • Dedicated transmitter can provide 1-10 mW at close range (< 1m)
  • Rectenna efficiency: 30-60%
  • Consider FCC/regulatory limits on transmitted power
  • Multiple frequency bands increase harvest opportunity

1619.6 Knowledge Check

CautionTest Your Understanding

Question 1: Your solar-powered sensor works fine in summer but fails in winter. What’s the most likely cause?

Show Answer

Underestimating seasonal variability. Winter months can have 70% less solar energy than summer due to:

  • Shorter days
  • Lower sun angle
  • More cloud cover

Solution: Redesign using worst-month (December) solar availability, or add secondary harvesting source.

Question 2: A supercapacitor-only system works during the day but fails overnight. What’s happening?

Show Answer

Supercapacitor self-discharge. Supercaps can lose 10-30% of charge per day through leakage.

Solutions:

  1. Add a small battery for overnight energy storage
  2. Use hybrid supercapacitor with lower leakage
  3. Reduce overnight power consumption further

Question 3: Why might a hybrid solar + thermal system be better than solar alone?

Show Answer

Multiple benefits:

  1. Redundancy - If clouds block solar, thermal still works
  2. Extended hours - Thermal works at night if ΔT exists
  3. Higher total energy - Sum of both sources
  4. Seasonal smoothing - Thermal may improve in winter (indoor heating creates ΔT)

1619.7 Summary

TipKey Takeaways
  1. Match source to application - Solar for outdoor, vibration for industrial, thermal for HVAC, RF for low-power supplementary
  2. Design for worst case - Account for seasonal variations, cloudy days, maintenance shutdowns
  3. Size storage appropriately - Balance autonomy requirements against cost and size constraints
  4. Use MPPT - Gains of 15-30% justify the added complexity
  5. Consider hybrid systems - Multiple sources improve reliability and energy availability
  6. Account for efficiency losses - Real-world systems are 60-80% efficient end-to-end
  7. Plan for lifecycle - Batteries degrade; supercapacitors offer longer cycle life

1619.8 Further Reading

1619.9 What’s Next

Now that you understand energy harvesting concepts and practical implementation: