HVAC Fault Detection Using BIM-Based Heat Flow Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Modern HVAC control systems are complex and lack effective means for Fault Detection and Diagnosis (FDD), leading to difficulties in fault localization and high engineering efforts due to the variety of HVAC systems, with existing methods being time-consuming and labor-intensive.

Innovation Solution

A Heat Flow Model (HFM) methodology that automatically translates Building Information Model (BIM) descriptions into HFM graphs, enabling direct integration into HVAC control systems for real-time fault detection and diagnosis, using a hierarchical model with nodes that simulate the dynamic physical behavior of HVAC components and propagate parameter ranges for fault detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If FDD systems are customized for specific HVAC systems, then fault detection accuracy is improved, but development effort and cost increase significantly

Engineering Contradiction:
Improvefault detection accuracyVSAvoiddevelopment effort
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent implements a universal FDD system based on BIM that can be applied to various HVAC systems without customization. The system uses standardized building information models and automated rule generation to achieve broad applicability across different building types and HVAC configurations, eliminating the need for system-specific development while maintaining effective fault detection

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent performs preliminary actions by automatically generating fault detection rules from BIM data during the design phase. The system pre-configures FDD capabilities using building information models, so that when the HVAC system is deployed, fault detection is already operational without requiring additional customization effort

Inventive Principle:
Principle #10Preliminary action

2Reliability

If manual rule derivation is performed for each HVAC system, then fault detection coverage is improved, but time consumption and labor intensity increase

Engineering Contradiction:
Improvefault detection coverageVSAvoidengineering efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent implements self-service by enabling the FDD system to automatically generate its own detection rules from BIM data without human intervention. The system extracts building information models, automatically creates fault detection rules based on HVAC system configurations, and configures itself, thereby eliminating manual rule derivation while maintaining comprehensive fault detection coverage

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces the manual mechanical process of rule derivation with an automated computational system. Instead of engineers manually creating fault detection rules, the system uses algorithms to automatically generate rules from BIM data, substituting human labor with automated processing while maintaining or improving detection coverage

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If ANN is trained offline with limited data, then fault classification capability is improved, but coverage of all fault types is reduced

Engineering Contradiction:
Improvefault classification capabilityVSAvoidfault type coverage
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent performs preliminary action by generating comprehensive fault detection rules from complete BIM data before the system operates. Instead of training ANN offline with limited historical data, the system pre-configures detection rules based on the full building information model, ensuring all potential fault types are covered from the start without relying on limited training datasets

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2491464B1Fault detection in HVAC-systems using building information models and heat flow models
Publication Date: 2016.09.14 SIEMENS CORP
  • EP2491464B1 patent drawingFigure 1
  • EP2491464B1 patent drawingFigure 2
  • EP2491464B1 patent drawingFigure 3

AI summary

Systems and methods are described that provide a Heat Flow Model (HFM) graph modeling methodology. Embodiments automatically translate formal HVAC system descriptions from a Building Information Model (BIM) into HFM graphs, and compile the graphs into executable FDD systems. During an engineering phase, a user interface is used to enter parameters, conditions, and switches not found in the BIM. During a runtime phase, real time data from an HVAC control system is input to the generated FDD system (HFM graph) for fault detection and diagnosis.