Building Entity Modeling for Economic Central Plant Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optimizing the allocation of energy loads across the assets of a central plant to minimize economic costs while meeting varying energy demands and real-time pricing conditions is challenging due to the complexity of managing multiple subplants and energy storage systems.
Innovation Solution
An economic model predictive control (EMPC) tool is implemented, which includes configuration tools for user input, a data model extender to define new entities and relationships, a high-level EMPC algorithm to generate optimization problems, and an asset allocator to determine optimal control decisions for operating the central plant, simulating operations over a predetermined time period, and presenting results to optimize resource production and consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If traditional control methods are used to operate central plant subplants, then operational simplicity is maintained, but economic costs increase due to inability to optimize resource allocation under real-time pricing conditions
Solution Approach 1:
The control system is segmented into multiple hierarchical levels: real-time control layer for immediate subplant operations, predictive control layer for optimization under real-time pricing, and scheduling layer for energy storage management. This segmentation allows complex optimization functions to be distributed across manageable modules, reducing overall system complexity while achieving economic cost reduction through coordinated operation of chiller plants, boiler plants, and energy storage systems
Solution Approach 2:
The system performs preliminary actions by predicting future energy prices and loads, then pre-scheduling energy storage charging/discharging operations and subplant operations in advance. The predictive control algorithm calculates optimal operation schedules before real-time execution, allowing the system to prepare optimization strategies ahead of time based on forecasted pricing conditions, thereby reducing actual economic costs without requiring complex real-time decision-making
2Adaptability or versatility
If energy storage systems are integrated into the central plant, then ability to shift resource consumption to lower-cost periods is improved, but system complexity increases due to additional coordination requirements
Solution Approach 1:
The energy storage systems are designed with multi-functionality to serve multiple purposes: peak shaving, load shifting, frequency regulation, and backup power supply. By making storage systems universal in their functionality, the patent reduces the need for specialized equipment for each function, thereby managing complexity while enhancing adaptability. The control system universally manages all storage assets across different subplants using a unified optimization framework
Solution Approach 2:
The predictive control algorithm acts as an intermediary between energy storage systems and real-time pricing signals. Rather than allowing direct complex interactions between multiple storage systems and varying price structures, the intermediary algorithm translates pricing signals into simplified charging/discharging schedules, mediating the complexity and enabling flexible resource consumption timing without proportionally increasing coordination complexity
3Productivity
If real-time optimization is implemented to minimize economic costs, then energy efficiency is improved, but computational requirements and control complexity increase
Solution Approach 1:
The optimization algorithm employs dynamic adjustment capabilities where control parameters and optimization horizons are adapted based on real-time conditions such as predicted price volatility, load variability, and storage state of charge. The system dynamically switches between different optimization strategies (e.g., aggressive charge/discharge during high-price periods vs. conservative operation during low-price periods), allowing efficient energy allocation without requiring excessively complex algorithms for all conditions simultaneously
Solution Approach 2:
The system changes key parameters such as prediction horizon length, optimization frequency, and constraint tightness based on real-time pricing conditions and system state. During periods of high price volatility, the algorithm extends prediction horizons and increases optimization frequency to capture more opportunities. During stable periods, it reduces computational intensity. This parameter adaptation enables high energy allocation efficiency while managing computational requirements through conditional complexity adjustment
Data Source
AI summary
Systems and methods for modeling and controlling entities of a building system are provided. An exemplary method includes comparing an input indicating a new entity or connection between entities of a building system to a data model for the building system to determine whether the new entity or connection is represented in the data model, extending the data model to define the new entity or connection in response to determining that the new entity or connection is not represented in the data model, and using the data model with the new entity or connection in a control strategy to generate control decisions for the entities of the building system.