ems

Energy Management System (EMS)

A comprehensive Go-based Energy Management System that optimizes energy consumption, production, and storage across multiple sources including solar (PV), battery storage, grid connection, and controllable loads. The system uses real-time electricity price data, weather forecasts, and Model Predictive Control (MPC) to minimize energy costs while maintaining system reliability.

Overview

The Energy Management System provides intelligent control and optimization for residential or small commercial energy systems. It integrates:

• Solar Power (PV): Real-time monitoring via Modbus, solar position calculations, and weather-based forecasting
• Battery Storage: Intelligent charge/discharge scheduling with degradation cost consideration
• Grid Connection: Dynamic import/export management based on real-time pricing
• Controllable Loads: Automated management of high-power devices (e.g., cryptocurrency miners, EV chargers, heat pumps)
• Price Optimization: Integration with ENTSO-E Transparency Platform for real-time electricity market prices
• Weather Integration: Meteorological data for solar production forecasting
• Web Dashboard: Real-time monitoring and control interface

Features

Energy Source Management

• Solar Power Monitoring: Real-time PV production tracking via Modbus (SigEnergy integration)
• Solar Forecasting: Weather-based production forecasting with sun position calculations
• Battery Management: SOC monitoring, intelligent charge/discharge scheduling, degradation tracking
• Grid Integration: Dynamic import/export based on price signals and system state

Load Management

• Automatic Device Discovery: Network scanning to find controllable loads (currently supports Avalon miners)
• Price-Based Scheduling: Automatically activates/deactivates loads based on configurable price thresholds
• Power-Based Control: Manages loads to stay within PV production limits or configured power budgets
• Thermal Protection: Monitors device temperatures and adjusts operation modes to prevent overheating
• Multiple Operating Modes: Support for eco, standard, and high-performance modes

Optimization & Control

• Model Predictive Control (MPC): Advanced optimization for battery scheduling and load management
• Real-time Price Monitoring: Fetches electricity prices from ENTSO-E API with caching
• Dynamic Scheduling: Adjusts system behavior based on price forecasts, weather, and system state
• Configurable Thresholds: Flexible configuration for prices, power limits, and operating modes

System Features

• Dry-run Mode: Test configurations without executing actual control commands
• Concurrent Operations: Manages multiple devices simultaneously for efficiency
• Robust Error Handling: Comprehensive retry logic and error recovery
• Health Monitoring: Optional health check endpoint for system monitoring
• Flexible Configuration: Support for both command-line arguments and configuration files
• Comprehensive Logging: Detailed logging with configurable levels and formats
• Web Interface: Real-time dashboard with WebSocket updates

Architecture

┌─────────────────────────────────────────────────────────────┐
│                  Energy Management System                    │
├─────────────────────────────────────────────────────────────┤
│                                                              │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────┐     │
│  │   Weather    │  │  ENTSO-E     │  │   Modbus     │     │
│  │  Forecasts   │  │  Pricing     │  │   (PV/Bat)   │     │
│  └──────┬───────┘  └──────┬───────┘  └──────┬───────┘     │
│         │                  │                  │              │
│         └──────────────────┼──────────────────┘              │
│                            ↓                                 │
│              ┌─────────────────────────┐                    │
│              │   MPC Optimization      │                    │
│              │   - Battery Schedule    │                    │
│              │   - Load Management     │                    │
│              │   - Cost Minimization   │                    │
│              └─────────┬───────────────┘                    │
│                        ↓                                     │
│         ┌──────────────────────────────┐                   │
│         │    Scheduler & Controller     │                   │
│         └──────────┬───────────────────┘                   │
│                    │                                         │
│        ┌───────────┼───────────┐                           │
│        ↓           ↓           ↓                            │
│   ┌────────┐  ┌────────┐  ┌────────┐                      │
│   │  PV    │  │Battery │  │ Loads  │                      │
│   │Monitor │  │Control │  │Control │                      │
│   └────────┘  └────────┘  └────────┘                      │
│                                                              │
│              ┌─────────────────────────┐                    │
│              │   Web Dashboard         │                    │
│              │   - Real-time Monitor   │                    │
│              │   - Manual Control      │                    │
│              │   - Charts & Analytics  │                    │
│              └─────────────────────────┘                    │
└─────────────────────────────────────────────────────────────┘

How It Works

The Energy Management System operates on multiple time scales:

Real-time Monitoring (10-second intervals, configurable)

1. PV Production: Reads current solar power production via Modbus
2. Battery State: Monitors State of Charge (SOC), power flow, and health
3. Grid Metrics: Tracks import/export power and cumulative energy
4. Data Logging: Stores metrics in PostgreSQL for analysis and visualization

Load State Management (1-minute intervals, configurable)

1. Device Monitoring: Checks state and health of all controllable loads
2. Temperature Protection: Monitors device temperatures and adjusts modes
   • High temperature: Switches to Eco mode to prevent overheating
   • Normal temperature: Returns to Standard/Super mode as configured
3. Power Limit Enforcement: Ensures total load stays within configured limits
4. Fault Detection: Identifies and logs device issues

Energy Optimization (15-minute intervals, configurable)

1. Price Updates: Fetches current and forecasted electricity prices
2. Weather Forecast: Updates solar production forecast
3. MPC Optimization: Calculates optimal battery and load schedules for next 24-48 hours
4. Schedule Execution: Implements optimized decisions for current period
5. Load Management:
   • Price below threshold: Activate loads, prioritize self-consumption
   • Price above threshold: Deactivate non-essential loads, maximize export
   • Excess PV available: Activate loads to consume surplus power
   • Limited PV: Deactivate loads or switch to economy modes

Thermal Protection

• High Temperature Detection: Fan speed monitoring (for miners) or temperature sensors
• Automatic Mode Reduction: Switches devices to lower power modes when overheating
• Gradual Recovery: Returns to higher performance modes when safe
• Configurable Thresholds: Customizable temperature/fan speed limits

Prerequisites

• Go 1.25.1 or later
• ENTSO-E Transparency Platform API token (optional, for price-based optimization)
• Network access to controllable devices
• Modbus-compatible PV/battery system (optional, for solar monitoring)
• PostgreSQL database (optional, for data logging)

Installation

From Source

git clone https://github.com/devskill-org/ems.git
cd energy-management-system
go build -o ems ./

Using Docker

docker build -t ems .
docker run ems

Configuration

Command Line Usage

# Basic usage with defaults
./ems

# Custom configuration
./ems --config custom.json

# Show plant information
./ems -info

# Run web server only (no automated control)
./ems -serverOnly

# Show help
./ems -help

Configuration File

Create a config.json file for system configuration:

{
  "price_limit": 50.0,
  "network": "192.168.1.0/24",
  "check_price_interval": "15m",
  "miner_discovery_interval": "10m",
  "miners_state_check_interval": "1m",
  "dry_run": false,
  "api_timeout": "30s",
  "log_level": "info",
  "log_format": "text",
  "miner_timeout": "5s",
  "health_check_port": 8080,
  "security_token": "<ENTSO-E API token>",
  "fanr_high_threshold": 70,
  "fanr_low_threshold": 50,
  "url_format": "https://web-api.tp.entsoe.eu/api?documentType=A44&out_Domain=10YLV-1001A00074&in_Domain=10YLV-1001A00074&periodStart=%s&periodEnd=%s&securityToken=%s",
  "location": "CET",
  "plant_modbus_address": "192.168.1.100:502",
  "device_id": 0,
  "pv_poll_interval": "10s",
  "pv_integration_period": "15m",
  "postgres_conn_string": "postgres://user:pass@host/dbname",
  "battery_capacity": 24.0,
  "battery_max_charge": 12.0,
  "battery_max_discharge": 12.0,
  "battery_min_soc": 0.0,
  "battery_max_soc": 1.0,
  "battery_efficiency": 0.92,
  "battery_degradation_cost": 0.05,
  "max_grid_import": 30.0,
  "max_grid_export": 30.0,
  "max_solar_power": 30.0,
  "import_price_operator_fee": 8.5,
  "import_price_delivery_fee": 40.0,
  "export_price_operator_fee": 17.0,
  "miners_power_limit": 30.0,
  "miner_power_standby": 0.05,
  "miner_power_eco": 0.8,
  "miner_power_standard": 1.6,
  "miner_power_super": 1.8,
  "use_pv_power_control": true,
  "latitude": 56.9496,
  "longitude": 24.1052,
  "weather_update_interval": "1h",
  "user_agent": "EMS/1.0 ([email protected])"
}

Configuration Options

Core Settings

Energy Sources

Battery Settings

Grid Settings

Load Management

Load Power Consumption

Weather & Location

Pricing API

Data Storage

Usage Examples

Basic Setup

1. Get your ENTSO-E API token from the transparency platform
2. Configure your system in config.json
3. Run the EMS:

   ./ems
   

Testing with Dry-run Mode

Test your configuration without controlling actual devices:

./ems --config config.json

Set "dry_run": true in your config file.

Example output:

[SCHEDULER] DRY-RUN MODE ENABLED: Actions will be simulated only
[SCHEDULER] Current electricity price: 65.50 EUR/MWh
[SCHEDULER] PV Production: 12.5 kW, Battery SOC: 85%
[SCHEDULER] Price limit: 60.00 EUR/MWh
[SCHEDULER] DRY-RUN: Would reduce load at 192.168.1.100:4028 (price too high)
[SCHEDULER] DRY-RUN: Would charge battery (excess PV available)

Monitoring System Status

Access the web dashboard at http://localhost:8080 (or your configured port) to:

• View real-time energy flow
• Monitor PV production and battery state
• See current electricity prices
• Control devices manually
• View historical data and charts

Viewing Plant Information

Check your PV/battery system configuration:

./ems -info

API Reference

Device States

The system recognizes multiple device states:

• Running (0): Device is active
• Mining (1): Device is actively processing (for miners)
• StandBy (2): Device is in standby/idle mode

Device Operating Modes

Controllable devices support multiple operating modes:

• Eco Mode (0): Reduced power consumption (e.g., 0.8 kW)
• Standard Mode (1): Normal operation (e.g., 1.6 kW)
• Super Mode (2): Maximum performance (e.g., 1.8 kW)

EMS Operations

• WakeUp: Activates device from standby
• Standby: Transitions device to standby state
• SetWorkMode: Changes device operating mode
• GetLiteStats: Retrieves current device status
• GetPlantRunningInfo: Retrieves complete plant running information via Modbus (PV power, battery SOC, grid power, ESS power, etc.)
• GetBatterySOC: Reads battery state of charge
• OptimizeSchedule: Runs MPC optimization

Web Dashboard

The integrated web interface provides:

• Real-time Monitoring: Live updates via WebSocket

  • Current PV production
  • Battery state and power flow
  • Grid import/export
  • Load status and power consumption

• Price Display: Current and forecasted electricity prices

• Device Control: Manual override of automatic control

  • Activate/deactivate loads
  • Change operating modes
  • Force battery charge/discharge

• Historical Data: Charts and graphs of system performance

  • Energy production and consumption
  • Price trends
  • Battery usage patterns

• System Status: Health monitoring and diagnostics

Access the dashboard at: http://<your-server-ip>:8080

Troubleshooting

Common Issues

1. No devices discovered

   • Check network configuration (CIDR notation)
   • Verify devices are accessible and have API enabled
   • Check firewall settings

2. Modbus connection errors

   • Verify plant_modbus_address is correct
   • Check network connectivity to PV system
   • Ensure Modbus TCP is enabled on the device

3. Price API errors

   • Verify security_token is correct
   • Check internet connectivity
   • Ensure ENTSO-E API is accessible
   • Verify url_format matches your region

4. Database connection issues

   • Verify PostgreSQL connection string
   • Check database permissions
   • Ensure database schema is initialized

5. Device control failures

   • Check device API port (default: 4028 for miners)
   • Verify device firmware supports required commands
   • Check network connectivity

Debug Mode

Enable detailed logging:

{
  "log_level": "debug",
  "log_format": "json"
}

Health Check

Monitor system health:

curl http://localhost:8080/health

Expected response:

{
  "status": "healthy",
  "timestamp": "2024-01-15T10:30:00Z",
  "pv_connected": true,
  "database_connected": true,
  "devices_discovered": 3
}

Use Cases

Residential Solar + Battery System

Optimize self-consumption and minimize grid import costs:

• Charge battery during low-price periods
• Discharge battery during high-price periods
• Activate high-power loads when excess PV is available
• Export surplus to grid when prices are favorable

Cryptocurrency Mining Operation

Maximize profitability by mining only when economical:

• Mine only when electricity prices are below threshold
• Scale mining power with available PV production
• Protect hardware with thermal management
• Track energy costs for profitability analysis

Small Commercial Energy Management

Optimize energy costs for small businesses:

• Schedule high-power equipment during low-price periods
• Maximize solar self-consumption
• Peak shaving with battery storage
• Demand response based on price signals

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

License

This project is licensed under the terms specified in the LICENSE file.

Acknowledgments

• ENTSO-E Transparency Platform for electricity price data
• Open-Meteo for weather forecasts
• SigEnergy for PV system integration
• The Go community for excellent libraries and tools

Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: Check the /docs folder for detailed guides

Roadmap

• Support for additional controllable load types (EV chargers, heat pumps, etc.)
• Machine learning for improved solar production forecasting
• Integration with additional battery systems
• Mobile app for remote monitoring
• Multi-site management
• Enhanced MPC algorithms
• Integration with home automation systems (Home Assistant, etc.)