Introduction
Build a comprehensive building energy monitoring system with CT sensors, power quality analysis, dashboard, and anomaly alerts. This comprehensive guide covers everything from design through implementation, testing, and deployment.
Building Energy Monitoring System
Build a comprehensive building energy monitoring system with CT sensors, power quality analysis, dashboard, and anomaly alerts.
DifficultyIntermediate
Duration3–4 weeks
Components10 items
Steps4 steps
Theory & Background
CT (Current Transformer) clamps around a live wire — measures current via electromagnetic induction (no electrical connection to live circuit). SCT-013-030: 30A input → 1V output (through 33Ω burden resistor). Calibrate: compare CT reading with a reference clamp meter on a known load (1kW heater). Calibration factor = reference_current / CT_raw_reading. AC current waveform is sampled at 10kHz by ADC, RMS calculated: I_rms = sqrt(mean(i²)). Sample at least 2 complete cycles (40ms at 50Hz) for accurate RMS.
Advertisement
Components & Requirements
10 components required for this project.
| # | Component | Purpose | Qty |
|---|---|---|---|
| 1 | Current Transformer (SCT-013-030, 30A) | Non-invasive current measurement | x6 |
| 2 | ZMPT101B Voltage Transformer Module | AC voltage measurement | x2 |
| 3 | Arduino Mega + ADS1115 ADC | High-resolution energy measurement | x1 |
| 4 | Raspberry Pi 4 | Data processing and IoT gateway | x1 |
| 5 | InfluxDB (time-series database) | Energy data storage | x1 |
| 6 | Grafana | Energy dashboard visualization | x1 |
| 7 | Telegram Bot API | Anomaly alerts on phone | x1 |
| 8 | RS-485 + Modbus (for digital meters) | Professional energy meter interfacing | x1 |
| 9 | DIN Rail enclosure | Panel-mounted professional installation | x1 |
| 10 | Circuit breakers (MCB) × 6 | Circuit protection | x6 |
Step-by-Step Implementation
Follow these 4 steps carefully.
1
CT Sensor Theory and Calibration
CT (Current Transformer) clamps around a live wire — measures current via electromagnetic induction (no electrical connection to live circuit). SCT-013-030: 30A input → 1V output (through 33Ω burden resistor). Calibrate: compare CT reading with a reference clamp meter on a known load (1kW heater). Calibration factor = reference_current / CT_raw_reading. AC current waveform is sampled at 10kHz by ADC, RMS calculated: I_rms = sqrt(mean(i²)). Sample at least 2 complete cycles (40ms at 50Hz) for accurate RMS.
2
Power Factor and Reactive Power
True power (W) = V_rms × I_rms × cos(φ) (power factor). Apparent power (VA) = V_rms × I_rms. Reactive power (VAr) = V_rms × I_rms × sin(φ). Power factor cos(φ): 1.0 = purely resistive (heater), < 1.0 = inductive (motors) or capacitive. Measure instantaneous power: p(t) = v(t) × i(t), integrate over cycle: P = mean(v×i). Compare P with |S| = V_rms × I_rms to get power factor directly without phase measurement. Poor power factor (< 0.8) incurs penalty charges from utilities.
3
Real-Time Dashboard and Analytics
Grafana panels: real-time power per circuit (W), energy today (kWh), power factor, total building power vs time. Derived analytics: load profile (power vs time of day), load duration curve (how long above various power levels), peak demand (maximum 15-minute average power — utility billing basis), standby power (overnight consumption when building unoccupied — target < 2% of peak). Cost calculation: kWh × tariff rate → daily/monthly cost estimate.
4
Anomaly Detection and Alerts
Rule-based alerts: circuit load > historical average by 30% → inspect. Overnight consumption > 500W (suspicious — HVAC leak, equipment left on) → alert. Power factor < 0.75 → notify for capacitor bank installation recommendation. Motor startup anomaly: normal startup current = 6× running current for < 3 seconds. If high current persists > 10 seconds → motor stuck or bearing failure. Machine learning anomaly detection: train isolation forest on 30 days of baseline data, alert when new day deviates significantly.
Code & Implementation
Core code for energy_monitor.ino:
energy_monitor.ino
C/C++
#include <Wire.h> #include <Adafruit_ADS1X15.h> Adafruit_ADS1115 ads; const float CAL_FACTOR = 30.0 / 1.0; // SCT-013-030: 30A per 1V const float VREF = 2.5; // Midpoint voltage (voltage divider) float measureRMS(int channel, int samples = 500) { float sum_sq = 0; for(int i = 0; i < samples; i++) { int16_t raw = ads.readADC_SingleEnded(channel); float voltage = (raw * 0.0001875) - VREF; // ADS1115 resolution: 187.5µV/bit at ±6.144V float current = voltage * CAL_FACTOR; sum_sq += current * current; delayMicroseconds(100); // 10kHz sampling } return sqrt(sum_sq / samples); } float measurePower(int v_channel, int i_channel, int samples = 500) { float sum_p = 0; for(int i = 0; i < samples; i++) { float v = (ads.readADC_SingleEnded(v_channel) * 0.0001875 - VREF) * 230; // Scale to 230V float cur = (ads.readADC_SingleEnded(i_channel) * 0.0001875 - VREF) * CAL_FACTOR; sum_p += v * cur; delayMicroseconds(100); } return sum_p / samples; // True power in Watts } void setup() { Serial.begin(115200); ads.setGain(GAIN_TWOTHIRDS); // ±6.144V range for ADS1115 ads.begin(); } void loop() { float I1 = measureRMS(0); // Circuit 1 current float I2 = measureRMS(1); // Circuit 2 current float P = measurePower(2, 0); // True power (circuit 1) float S = 230 * I1; // Apparent power float PF = abs(P) / S; // Power factor // Transmit as JSON to Raspberry Pi Serial.print("{\"I1\":"); Serial.print(I1, 2); Serial.print(",\"I2\":"); Serial.print(I2, 2); Serial.print(",\"P\":"); Serial.print(P, 1); Serial.print(",\"PF\":"); Serial.print(PF, 3); Serial.println("}"); delay(1000); }Testing & Troubleshooting
Test Building Energy Monitoring System by verifying each subsystem individually before full integration.
Troubleshooting Tips
Verify power voltages, check ground connections, use serial monitor for debug.
Real-World Applications
Commercial building energy management
Industrial facility energy audit
Hospital critical power monitoring
Data center power usage effectiveness (PUE)
Smart home energy awareness
Utility demand response program
Solar-grid energy balance monitoring
EV charging load management
Extensions & Next Steps
- Integrate with solar generation monitoring for net energy tracking
- Add weather correlation to separate weather from behavioral energy changes
- Build automated load shedding for peak demand management
- Implement digital twin of building energy model for optimization
- Add thermal comfort sensors to correlate HVAC usage with comfort
Interactive Playground
Coming SoonAn interactive simulator will be available here — simulate circuits and run code in-browser without hardware.
Frequently Asked Questions
How much energy can a building monitoring system typically save?
Building energy monitoring without action: 0% savings. Monitoring with awareness and manual action: 5–15% reduction (identify wasted standby power, peak demand reduction, equipment scheduling). Automated demand response integration: 10–20% reduction (automatically reduce non-critical loads during peak tariff hours). Combined monitoring + optimization + occupancy integration: 20–35% reduction commonly reported in commercial buildings. ROI: monitoring equipment cost Rs 20,000–50,000 for a commercial building, annual savings Rs 50,000–200,000 → payback 3–6 months.
Advertisement